Upload 11 files

.dockerignore

@@ -0,0 +1,9 @@
+.git
+__pycache__
+*.pyc
+.vscode
+.idea
+*.md
+Dockerfile
+docker-compose.yml
+.dockerignore

Dockerfile

@@ -0,0 +1,20 @@
+FROM python:3.10-slim
+
+WORKDIR /app
+
+COPY requirements.txt .
+
+RUN pip install --no-cache-dir -r requirements.txt
+
+COPY . .
+
+EXPOSE 8000
+
+ENV PYTHONUNBUFFERED=1 \
+    PYTHONIOENCODING=UTF-8 \
+    PYTHONDONTWRITEBYTECODE=1 \
+    CONFIG_PATH=/app/config.yaml
+
+RUN chmod +x /app/docker-entrypoint.sh
+
+ENTRYPOINT ["/app/docker-entrypoint.sh"]

LICENSE

@@ -0,0 +1,674 @@
+                    GNU GENERAL PUBLIC LICENSE
+                       Version 3, 29 June 2007
+
+ Copyright (C) 2007 Free Software Foundation, Inc. <https://fsf.org/>
+ Everyone is permitted to copy and distribute verbatim copies
+ of this license document, but changing it is not allowed.
+
+                            Preamble
+
+  The GNU General Public License is a free, copyleft license for
+software and other kinds of works.
+
+  The licenses for most software and other practical works are designed
+to take away your freedom to share and change the works.  By contrast,
+the GNU General Public License is intended to guarantee your freedom to
+share and change all versions of a program--to make sure it remains free
+software for all its users.  We, the Free Software Foundation, use the
+GNU General Public License for most of our software; it applies also to
+any other work released this way by its authors.  You can apply it to
+your programs, too.
+
+  When we speak of free software, we are referring to freedom, not
+price.  Our General Public Licenses are designed to make sure that you
+have the freedom to distribute copies of free software (and charge for
+them if you wish), that you receive source code or can get it if you
+want it, that you can change the software or use pieces of it in new
+free programs, and that you know you can do these things.
+
+  To protect your rights, we need to prevent others from denying you
+these rights or asking you to surrender the rights.  Therefore, you have
+certain responsibilities if you distribute copies of the software, or if
+you modify it: responsibilities to respect the freedom of others.
+
+  For example, if you distribute copies of such a program, whether
+gratis or for a fee, you must pass on to the recipients the same
+freedoms that you received.  You must make sure that they, too, receive
+or can get the source code.  And you must show them these terms so they
+know their rights.
+
+  Developers that use the GNU GPL protect your rights with two steps:
+(1) assert copyright on the software, and (2) offer you this License
+giving you legal permission to copy, distribute and/or modify it.
+
+  For the developers' and authors' protection, the GPL clearly explains
+that there is no warranty for this free software.  For both users' and
+authors' sake, the GPL requires that modified versions be marked as
+changed, so that their problems will not be attributed erroneously to
+authors of previous versions.
+
+  Some devices are designed to deny users access to install or run
+modified versions of the software inside them, although the manufacturer
+can do so.  This is fundamentally incompatible with the aim of
+protecting users' freedom to change the software.  The systematic
+pattern of such abuse occurs in the area of products for individuals to
+use, which is precisely where it is most unacceptable.  Therefore, we
+have designed this version of the GPL to prohibit the practice for those
+products.  If such problems arise substantially in other domains, we
+stand ready to extend this provision to those domains in future versions
+of the GPL, as needed to protect the freedom of users.
+
+  Finally, every program is threatened constantly by software patents.
+States should not allow patents to restrict development and use of
+software on general-purpose computers, but in those that do, we wish to
+avoid the special danger that patents applied to a free program could
+make it effectively proprietary.  To prevent this, the GPL assures that
+patents cannot be used to render the program non-free.
+
+  The precise terms and conditions for copying, distribution and
+modification follow.
+
+                       TERMS AND CONDITIONS
+
+  0. Definitions.
+
+  "This License" refers to version 3 of the GNU General Public License.
+
+  "Copyright" also means copyright-like laws that apply to other kinds of
+works, such as semiconductor masks.
+
+  "The Program" refers to any copyrightable work licensed under this
+License.  Each licensee is addressed as "you".  "Licensees" and
+"recipients" may be individuals or organizations.
+
+  To "modify" a work means to copy from or adapt all or part of the work
+in a fashion requiring copyright permission, other than the making of an
+exact copy.  The resulting work is called a "modified version" of the
+earlier work or a work "based on" the earlier work.
+
+  A "covered work" means either the unmodified Program or a work based
+on the Program.
+
+  To "propagate" a work means to do anything with it that, without
+permission, would make you directly or secondarily liable for
+infringement under applicable copyright law, except executing it on a
+computer or modifying a private copy.  Propagation includes copying,
+distribution (with or without modification), making available to the
+public, and in some countries other activities as well.
+
+  To "convey" a work means any kind of propagation that enables other
+parties to make or receive copies.  Mere interaction with a user through
+a computer network, with no transfer of a copy, is not conveying.
+
+  An interactive user interface displays "Appropriate Legal Notices"
+to the extent that it includes a convenient and prominently visible
+feature that (1) displays an appropriate copyright notice, and (2)
+tells the user that there is no warranty for the work (except to the
+extent that warranties are provided), that licensees may convey the
+work under this License, and how to view a copy of this License.  If
+the interface presents a list of user commands or options, such as a
+menu, a prominent item in the list meets this criterion.
+
+  1. Source Code.
+
+  The "source code" for a work means the preferred form of the work
+for making modifications to it.  "Object code" means any non-source
+form of a work.
+
+  A "Standard Interface" means an interface that either is an official
+standard defined by a recognized standards body, or, in the case of
+interfaces specified for a particular programming language, one that
+is widely used among developers working in that language.
+
+  The "System Libraries" of an executable work include anything, other
+than the work as a whole, that (a) is included in the normal form of
+packaging a Major Component, but which is not part of that Major
+Component, and (b) serves only to enable use of the work with that
+Major Component, or to implement a Standard Interface for which an
+implementation is available to the public in source code form.  A
+"Major Component", in this context, means a major essential component
+(kernel, window system, and so on) of the specific operating system
+(if any) on which the executable work runs, or a compiler used to
+produce the work, or an object code interpreter used to run it.
+
+  The "Corresponding Source" for a work in object code form means all
+the source code needed to generate, install, and (for an executable
+work) run the object code and to modify the work, including scripts to
+control those activities.  However, it does not include the work's
+System Libraries, or general-purpose tools or generally available free
+programs which are used unmodified in performing those activities but
+which are not part of the work.  For example, Corresponding Source
+includes interface definition files associated with source files for
+the work, and the source code for shared libraries and dynamically
+linked subprograms that the work is specifically designed to require,
+such as by intimate data communication or control flow between those
+subprograms and other parts of the work.
+
+  The Corresponding Source need not include anything that users
+can regenerate automatically from other parts of the Corresponding
+Source.
+
+  The Corresponding Source for a work in source code form is that
+same work.
+
+  2. Basic Permissions.
+
+  All rights granted under this License are granted for the term of
+copyright on the Program, and are irrevocable provided the stated
+conditions are met.  This License explicitly affirms your unlimited
+permission to run the unmodified Program.  The output from running a
+covered work is covered by this License only if the output, given its
+content, constitutes a covered work.  This License acknowledges your
+rights of fair use or other equivalent, as provided by copyright law.
+
+  You may make, run and propagate covered works that you do not
+convey, without conditions so long as your license otherwise remains
+in force.  You may convey covered works to others for the sole purpose
+of having them make modifications exclusively for you, or provide you
+with facilities for running those works, provided that you comply with
+the terms of this License in conveying all material for which you do
+not control copyright.  Those thus making or running the covered works
+for you must do so exclusively on your behalf, under your direction
+and control, on terms that prohibit them from making any copies of
+your copyrighted material outside their relationship with you.
+
+  Conveying under any other circumstances is permitted solely under
+the conditions stated below.  Sublicensing is not allowed; section 10
+makes it unnecessary.
+
+  3. Protecting Users' Legal Rights From Anti-Circumvention Law.
+
+  No covered work shall be deemed part of an effective technological
+measure under any applicable law fulfilling obligations under article
+11 of the WIPO copyright treaty adopted on 20 December 1996, or
+similar laws prohibiting or restricting circumvention of such
+measures.
+
+  When you convey a covered work, you waive any legal power to forbid
+circumvention of technological measures to the extent such circumvention
+is effected by exercising rights under this License with respect to
+the covered work, and you disclaim any intention to limit operation or
+modification of the work as a means of enforcing, against the work's
+users, your or third parties' legal rights to forbid circumvention of
+technological measures.
+
+  4. Conveying Verbatim Copies.
+
+  You may convey verbatim copies of the Program's source code as you
+receive it, in any medium, provided that you conspicuously and
+appropriately publish on each copy an appropriate copyright notice;
+keep intact all notices stating that this License and any
+non-permissive terms added in accord with section 7 apply to the code;
+keep intact all notices of the absence of any warranty; and give all
+recipients a copy of this License along with the Program.
+
+  You may charge any price or no price for each copy that you convey,
+and you may offer support or warranty protection for a fee.
+
+  5. Conveying Modified Source Versions.
+
+  You may convey a work based on the Program, or the modifications to
+produce it from the Program, in the form of source code under the
+terms of section 4, provided that you also meet all of these conditions:
+
+    a) The work must carry prominent notices stating that you modified
+    it, and giving a relevant date.
+
+    b) The work must carry prominent notices stating that it is
+    released under this License and any conditions added under section
+    7.  This requirement modifies the requirement in section 4 to
+    "keep intact all notices".
+
+    c) You must license the entire work, as a whole, under this
+    License to anyone who comes into possession of a copy.  This
+    License will therefore apply, along with any applicable section 7
+    additional terms, to the whole of the work, and all its parts,
+    regardless of how they are packaged.  This License gives no
+    permission to license the work in any other way, but it does not
+    invalidate such permission if you have separately received it.
+
+    d) If the work has interactive user interfaces, each must display
+    Appropriate Legal Notices; however, if the Program has interactive
+    interfaces that do not display Appropriate Legal Notices, your
+    work need not make them do so.
+
+  A compilation of a covered work with other separate and independent
+works, which are not by their nature extensions of the covered work,
+and which are not combined with it such as to form a larger program,
+in or on a volume of a storage or distribution medium, is called an
+"aggregate" if the compilation and its resulting copyright are not
+used to limit the access or legal rights of the compilation's users
+beyond what the individual works permit.  Inclusion of a covered work
+in an aggregate does not cause this License to apply to the other
+parts of the aggregate.
+
+  6. Conveying Non-Source Forms.
+
+  You may convey a covered work in object code form under the terms
+of sections 4 and 5, provided that you also convey the
+machine-readable Corresponding Source under the terms of this License,
+in one of these ways:
+
+    a) Convey the object code in, or embodied in, a physical product
+    (including a physical distribution medium), accompanied by the
+    Corresponding Source fixed on a durable physical medium
+    customarily used for software interchange.
+
+    b) Convey the object code in, or embodied in, a physical product
+    (including a physical distribution medium), accompanied by a
+    written offer, valid for at least three years and valid for as
+    long as you offer spare parts or customer support for that product
+    model, to give anyone who possesses the object code either (1) a
+    copy of the Corresponding Source for all the software in the
+    product that is covered by this License, on a durable physical
+    medium customarily used for software interchange, for a price no
+    more than your reasonable cost of physically performing this
+    conveying of source, or (2) access to copy the
+    Corresponding Source from a network server at no charge.
+
+    c) Convey individual copies of the object code with a copy of the
+    written offer to provide the Corresponding Source.  This
+    alternative is allowed only occasionally and noncommercially, and
+    only if you received the object code with such an offer, in accord
+    with subsection 6b.
+
+    d) Convey the object code by offering access from a designated
+    place (gratis or for a charge), and offer equivalent access to the
+    Corresponding Source in the same way through the same place at no
+    further charge.  You need not require recipients to copy the
+    Corresponding Source along with the object code.  If the place to
+    copy the object code is a network server, the Corresponding Source
+    may be on a different server (operated by you or a third party)
+    that supports equivalent copying facilities, provided you maintain
+    clear directions next to the object code saying where to find the
+    Corresponding Source.  Regardless of what server hosts the
+    Corresponding Source, you remain obligated to ensure that it is
+    available for as long as needed to satisfy these requirements.
+
+    e) Convey the object code using peer-to-peer transmission, provided
+    you inform other peers where the object code and Corresponding
+    Source of the work are being offered to the general public at no
+    charge under subsection 6d.
+
+  A separable portion of the object code, whose source code is excluded
+from the Corresponding Source as a System Library, need not be
+included in conveying the object code work.
+
+  A "User Product" is either (1) a "consumer product", which means any
+tangible personal property which is normally used for personal, family,
+or household purposes, or (2) anything designed or sold for incorporation
+into a dwelling.  In determining whether a product is a consumer product,
+doubtful cases shall be resolved in favor of coverage.  For a particular
+product received by a particular user, "normally used" refers to a
+typical or common use of that class of product, regardless of the status
+of the particular user or of the way in which the particular user
+actually uses, or expects or is expected to use, the product.  A product
+is a consumer product regardless of whether the product has substantial
+commercial, industrial or non-consumer uses, unless such uses represent
+the only significant mode of use of the product.
+
+  "Installation Information" for a User Product means any methods,
+procedures, authorization keys, or other information required to install
+and execute modified versions of a covered work in that User Product from
+a modified version of its Corresponding Source.  The information must
+suffice to ensure that the continued functioning of the modified object
+code is in no case prevented or interfered with solely because
+modification has been made.
+
+  If you convey an object code work under this section in, or with, or
+specifically for use in, a User Product, and the conveying occurs as
+part of a transaction in which the right of possession and use of the
+User Product is transferred to the recipient in perpetuity or for a
+fixed term (regardless of how the transaction is characterized), the
+Corresponding Source conveyed under this section must be accompanied
+by the Installation Information.  But this requirement does not apply
+if neither you nor any third party retains the ability to install
+modified object code on the User Product (for example, the work has
+been installed in ROM).
+
+  The requirement to provide Installation Information does not include a
+requirement to continue to provide support service, warranty, or updates
+for a work that has been modified or installed by the recipient, or for
+the User Product in which it has been modified or installed.  Access to a
+network may be denied when the modification itself materially and
+adversely affects the operation of the network or violates the rules and
+protocols for communication across the network.
+
+  Corresponding Source conveyed, and Installation Information provided,
+in accord with this section must be in a format that is publicly
+documented (and with an implementation available to the public in
+source code form), and must require no special password or key for
+unpacking, reading or copying.
+
+  7. Additional Terms.
+
+  "Additional permissions" are terms that supplement the terms of this
+License by making exceptions from one or more of its conditions.
+Additional permissions that are applicable to the entire Program shall
+be treated as though they were included in this License, to the extent
+that they are valid under applicable law.  If additional permissions
+apply only to part of the Program, that part may be used separately
+under those permissions, but the entire Program remains governed by
+this License without regard to the additional permissions.
+
+  When you convey a copy of a covered work, you may at your option
+remove any additional permissions from that copy, or from any part of
+it.  (Additional permissions may be written to require their own
+removal in certain cases when you modify the work.)  You may place
+additional permissions on material, added by you to a covered work,
+for which you have or can give appropriate copyright permission.
+
+  Notwithstanding any other provision of this License, for material you
+add to a covered work, you may (if authorized by the copyright holders of
+that material) supplement the terms of this License with terms:
+
+    a) Disclaiming warranty or limiting liability differently from the
+    terms of sections 15 and 16 of this License; or
+
+    b) Requiring preservation of specified reasonable legal notices or
+    author attributions in that material or in the Appropriate Legal
+    Notices displayed by works containing it; or
+
+    c) Prohibiting misrepresentation of the origin of that material, or
+    requiring that modified versions of such material be marked in
+    reasonable ways as different from the original version; or
+
+    d) Limiting the use for publicity purposes of names of licensors or
+    authors of the material; or
+
+    e) Declining to grant rights under trademark law for use of some
+    trade names, trademarks, or service marks; or
+
+    f) Requiring indemnification of licensors and authors of that
+    material by anyone who conveys the material (or modified versions of
+    it) with contractual assumptions of liability to the recipient, for
+    any liability that these contractual assumptions directly impose on
+    those licensors and authors.
+
+  All other non-permissive additional terms are considered "further
+restrictions" within the meaning of section 10.  If the Program as you
+received it, or any part of it, contains a notice stating that it is
+governed by this License along with a term that is a further
+restriction, you may remove that term.  If a license document contains
+a further restriction but permits relicensing or conveying under this
+License, you may add to a covered work material governed by the terms
+of that license document, provided that the further restriction does
+not survive such relicensing or conveying.
+
+  If you add terms to a covered work in accord with this section, you
+must place, in the relevant source files, a statement of the
+additional terms that apply to those files, or a notice indicating
+where to find the applicable terms.
+
+  Additional terms, permissive or non-permissive, may be stated in the
+form of a separately written license, or stated as exceptions;
+the above requirements apply either way.
+
+  8. Termination.
+
+  You may not propagate or modify a covered work except as expressly
+provided under this License.  Any attempt otherwise to propagate or
+modify it is void, and will automatically terminate your rights under
+this License (including any patent licenses granted under the third
+paragraph of section 11).
+
+  However, if you cease all violation of this License, then your
+license from a particular copyright holder is reinstated (a)
+provisionally, unless and until the copyright holder explicitly and
+finally terminates your license, and (b) permanently, if the copyright
+holder fails to notify you of the violation by some reasonable means
+prior to 60 days after the cessation.
+
+  Moreover, your license from a particular copyright holder is
+reinstated permanently if the copyright holder notifies you of the
+violation by some reasonable means, this is the first time you have
+received notice of violation of this License (for any work) from that
+copyright holder, and you cure the violation prior to 30 days after
+your receipt of the notice.
+
+  Termination of your rights under this section does not terminate the
+licenses of parties who have received copies or rights from you under
+this License.  If your rights have been terminated and not permanently
+reinstated, you do not qualify to receive new licenses for the same
+material under section 10.
+
+  9. Acceptance Not Required for Having Copies.
+
+  You are not required to accept this License in order to receive or
+run a copy of the Program.  Ancillary propagation of a covered work
+occurring solely as a consequence of using peer-to-peer transmission
+to receive a copy likewise does not require acceptance.  However,
+nothing other than this License grants you permission to propagate or
+modify any covered work.  These actions infringe copyright if you do
+not accept this License.  Therefore, by modifying or propagating a
+covered work, you indicate your acceptance of this License to do so.
+
+  10. Automatic Licensing of Downstream Recipients.
+
+  Each time you convey a covered work, the recipient automatically
+receives a license from the original licensors, to run, modify and
+propagate that work, subject to this License.  You are not responsible
+for enforcing compliance by third parties with this License.
+
+  An "entity transaction" is a transaction transferring control of an
+organization, or substantially all assets of one, or subdividing an
+organization, or merging organizations.  If propagation of a covered
+work results from an entity transaction, each party to that
+transaction who receives a copy of the work also receives whatever
+licenses to the work the party's predecessor in interest had or could
+give under the previous paragraph, plus a right to possession of the
+Corresponding Source of the work from the predecessor in interest, if
+the predecessor has it or can get it with reasonable efforts.
+
+  You may not impose any further restrictions on the exercise of the
+rights granted or affirmed under this License.  For example, you may
+not impose a license fee, royalty, or other charge for exercise of
+rights granted under this License, and you may not initiate litigation
+(including a cross-claim or counterclaim in a lawsuit) alleging that
+any patent claim is infringed by making, using, selling, offering for
+sale, or importing the Program or any portion of it.
+
+  11. Patents.
+
+  A "contributor" is a copyright holder who authorizes use under this
+License of the Program or a work on which the Program is based.  The
+work thus licensed is called the contributor's "contributor version".
+
+  A contributor's "essential patent claims" are all patent claims
+owned or controlled by the contributor, whether already acquired or
+hereafter acquired, that would be infringed by some manner, permitted
+by this License, of making, using, or selling its contributor version,
+but do not include claims that would be infringed only as a
+consequence of further modification of the contributor version.  For
+purposes of this definition, "control" includes the right to grant
+patent sublicenses in a manner consistent with the requirements of
+this License.
+
+  Each contributor grants you a non-exclusive, worldwide, royalty-free
+patent license under the contributor's essential patent claims, to
+make, use, sell, offer for sale, import and otherwise run, modify and
+propagate the contents of its contributor version.
+
+  In the following three paragraphs, a "patent license" is any express
+agreement or commitment, however denominated, not to enforce a patent
+(such as an express permission to practice a patent or covenant not to
+sue for patent infringement).  To "grant" such a patent license to a
+party means to make such an agreement or commitment not to enforce a
+patent against the party.
+
+  If you convey a covered work, knowingly relying on a patent license,
+and the Corresponding Source of the work is not available for anyone
+to copy, free of charge and under the terms of this License, through a
+publicly available network server or other readily accessible means,
+then you must either (1) cause the Corresponding Source to be so
+available, or (2) arrange to deprive yourself of the benefit of the
+patent license for this particular work, or (3) arrange, in a manner
+consistent with the requirements of this License, to extend the patent
+license to downstream recipients.  "Knowingly relying" means you have
+actual knowledge that, but for the patent license, your conveying the
+covered work in a country, or your recipient's use of the covered work
+in a country, would infringe one or more identifiable patents in that
+country that you have reason to believe are valid.
+
+  If, pursuant to or in connection with a single transaction or
+arrangement, you convey, or propagate by procuring conveyance of, a
+covered work, and grant a patent license to some of the parties
+receiving the covered work authorizing them to use, propagate, modify
+or convey a specific copy of the covered work, then the patent license
+you grant is automatically extended to all recipients of the covered
+work and works based on it.
+
+  A patent license is "discriminatory" if it does not include within
+the scope of its coverage, prohibits the exercise of, or is
+conditioned on the non-exercise of one or more of the rights that are
+specifically granted under this License.  You may not convey a covered
+work if you are a party to an arrangement with a third party that is
+in the business of distributing software, under which you make payment
+to the third party based on the extent of your activity of conveying
+the work, and under which the third party grants, to any of the
+parties who would receive the covered work from you, a discriminatory
+patent license (a) in connection with copies of the covered work
+conveyed by you (or copies made from those copies), or (b) primarily
+for and in connection with specific products or compilations that
+contain the covered work, unless you entered into that arrangement,
+or that patent license was granted, prior to 28 March 2007.
+
+  Nothing in this License shall be construed as excluding or limiting
+any implied license or other defenses to infringement that may
+otherwise be available to you under applicable patent law.
+
+  12. No Surrender of Others' Freedom.
+
+  If conditions are imposed on you (whether by court order, agreement or
+otherwise) that contradict the conditions of this License, they do not
+excuse you from the conditions of this License.  If you cannot convey a
+covered work so as to satisfy simultaneously your obligations under this
+License and any other pertinent obligations, then as a consequence you may
+not convey it at all.  For example, if you agree to terms that obligate you
+to collect a royalty for further conveying from those to whom you convey
+the Program, the only way you could satisfy both those terms and this
+License would be to refrain entirely from conveying the Program.
+
+  13. Use with the GNU Affero General Public License.
+
+  Notwithstanding any other provision of this License, you have
+permission to link or combine any covered work with a work licensed
+under version 3 of the GNU Affero General Public License into a single
+combined work, and to convey the resulting work.  The terms of this
+License will continue to apply to the part which is the covered work,
+but the special requirements of the GNU Affero General Public License,
+section 13, concerning interaction through a network will apply to the
+combination as such.
+
+  14. Revised Versions of this License.
+
+  The Free Software Foundation may publish revised and/or new versions of
+the GNU General Public License from time to time.  Such new versions will
+be similar in spirit to the present version, but may differ in detail to
+address new problems or concerns.
+
+  Each version is given a distinguishing version number.  If the
+Program specifies that a certain numbered version of the GNU General
+Public License "or any later version" applies to it, you have the
+option of following the terms and conditions either of that numbered
+version or of any later version published by the Free Software
+Foundation.  If the Program does not specify a version number of the
+GNU General Public License, you may choose any version ever published
+by the Free Software Foundation.
+
+  If the Program specifies that a proxy can decide which future
+versions of the GNU General Public License can be used, that proxy's
+public statement of acceptance of a version permanently authorizes you
+to choose that version for the Program.
+
+  Later license versions may give you additional or different
+permissions.  However, no additional obligations are imposed on any
+author or copyright holder as a result of your choosing to follow a
+later version.
+
+  15. Disclaimer of Warranty.
+
+  THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY
+APPLICABLE LAW.  EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT
+HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY
+OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO,
+THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+PURPOSE.  THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM
+IS WITH YOU.  SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF
+ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
+
+  16. Limitation of Liability.
+
+  IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
+WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS
+THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY
+GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE
+USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF
+DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD
+PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS),
+EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF
+SUCH DAMAGES.
+
+  17. Interpretation of Sections 15 and 16.
+
+  If the disclaimer of warranty and limitation of liability provided
+above cannot be given local legal effect according to their terms,
+reviewing courts shall apply local law that most closely approximates
+an absolute waiver of all civil liability in connection with the
+Program, unless a warranty or assumption of liability accompanies a
+copy of the Program in return for a fee.
+
+                     END OF TERMS AND CONDITIONS
+
+            How to Apply These Terms to Your New Programs
+
+  If you develop a new program, and you want it to be of the greatest
+possible use to the public, the best way to achieve this is to make it
+free software which everyone can redistribute and change under these terms.
+
+  To do so, attach the following notices to the program.  It is safest
+to attach them to the start of each source file to most effectively
+state the exclusion of warranty; and each file should have at least
+the "copyright" line and a pointer to where the full notice is found.
+
+    <one line to give the program's name and a brief idea of what it does.>
+    Copyright (C) <year>  <name of author>
+
+    This program is free software: you can redistribute it and/or modify
+    it under the terms of the GNU General Public License as published by
+    the Free Software Foundation, either version 3 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU General Public License for more details.
+
+    You should have received a copy of the GNU General Public License
+    along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+Also add information on how to contact you by electronic and paper mail.
+
+  If the program does terminal interaction, make it output a short
+notice like this when it starts in an interactive mode:
+
+    <program>  Copyright (C) <year>  <name of author>
+    This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
+    This is free software, and you are welcome to redistribute it
+    under certain conditions; type `show c' for details.
+
+The hypothetical commands `show w' and `show c' should show the appropriate
+parts of the General Public License.  Of course, your program's commands
+might be different; for a GUI interface, you would use an "about box".
+
+  You should also get your employer (if you work as a programmer) or school,
+if any, to sign a "copyright disclaimer" for the program, if necessary.
+For more information on this, and how to apply and follow the GNU GPL, see
+<https://www.gnu.org/licenses/>.
+
+  The GNU General Public License does not permit incorporating your program
+into proprietary programs.  If your program is a subroutine library, you
+may consider it more useful to permit linking proprietary applications with
+the library.  If this is what you want to do, use the GNU Lesser General
+Public License instead of this License.  But first, please read
+<https://www.gnu.org/licenses/why-not-lgpl.html>.

config.example.yaml

@@ -0,0 +1,93 @@
+# Toolify Configuration Example File
+# Please copy this file as config.yaml and modify the configuration according to your actual needs
+
+# Server configuration
+server:
+  port: 8000                    # Server listening port
+  host: "0.0.0.0"              # Server listening address
+  timeout: 180                  # Request timeout (seconds)
+
+# Upstream OpenAI compatible service configuration
+upstream_services:
+  - name: "openai"
+    base_url: "https://api.openai.com/v1"
+    api_key: "your-openai-api-key-here"
+    description: "OpenAI Official Service"
+    is_default: true
+    models:
+      - "gpt-3.5-turbo"
+      - "gpt-3.5-turbo-16k"
+      - "gpt-4"
+      - "gpt-4-turbo"
+      - "gpt-4o"
+      - "gpt-4o-mini"
+
+  - name: "google"
+    base_url: "https://generativelanguage.googleapis.com/v1beta/openai/"
+    api_key: "your-google-api-key-here"
+    description: "Google Gemini Service"
+    is_default: false
+    models:
+      # Use alias "gemini-2.5" to randomly select one of the following models
+      - "gemini-2.5:gemini-2.5-pro"
+      - "gemini-2.5:gemini-2.5-flash"
+      # You can also define models that can be used directly
+      - "gemini-2.5-pro"
+      - "gemini-2.5-flash"
+
+# Client authentication configuration
+client_authentication:
+  allowed_keys:
+    - "sk-my-secret-key-1"
+    - "sk-my-secret-key-2"
+
+# Feature configuration
+features:
+  enable_function_calling: true  # Enable function calling feature
+  log_level: "INFO"              # Logging level: DEBUG, INFO, WARNING, ERROR, CRITICAL, or DISABLED
+  convert_developer_to_system: true  # Whether to convert the developer role to the system role
+  key_passthrough: false          # If true, directly forward client-provided API key to upstream instead of using configured upstream key
+  model_passthrough: false         # If true, forward all requests directly to the 'openai' upstream service, ignoring model-based routing
+  
+  # Custom prompt template (optional). If not provided, the default prompt will be used.
+  # The default prompt includes comprehensive features:
+  # - Support for multiple tool calls in a single response
+  # - Context awareness to avoid duplicate tool calls
+  # - Strict parameter matching rules (preserving special characters like hyphens)
+  # - Clear format requirements with correct and incorrect examples
+  # - Tool result tracking via XML tags
+  #
+  # You can uncomment and customize the following template if needed:
+  # prompt_template: |
+  #   Your custom prompt template here...
+  #   Must include {tools_list} and {trigger_signal} placeholders
+
+# Configuration explanation:
+# 1. upstream_services: Configure multiple OpenAI compatible API services
+#    - name: Service name (for identification)
+#    - base_url: Base URL of the service
+#    - api_key: API key for the corresponding service
+#    - models: Complete list of models supported by the service
+#    - is_default: Whether it is the default service (used when the requested model is not in any service's model list)
+#    - description: Service description (optional)
+#
+# 2. Routing matching rules:
+#    - The system will exactly match the corresponding service based on the model name in the request
+#    - If the model name is not in the models list of any service, the service with is_default set to true will be used
+#    - There must be one and only one service marked as is_default: true
+#
+# 3. Client authentication:
+#    - allowed_keys: List of client API keys allowed to access this middleware
+#
+# 4. Logging levels:
+#    - DEBUG: Show all debug information (most verbose)
+#    - INFO: Show general information, warnings and errors
+#    - WARNING: Show only warnings and errors
+#    - ERROR: Show only errors
+#    - CRITICAL: Show only critical errors
+#    - DISABLED: Disable all logging
+#
+# 5. Security reminders:
+#    - Please keep API keys safe and do not commit configuration files containing real keys to version control systems
+#    - It is recommended to use different configuration files for different environments
+#    - Environment variables can be used to manage sensitive information

config_loader.py

@@ -0,0 +1,230 @@
+# SPDX-License-Identifier: GPL-3.0-or-later
+#
+# Toolify: Empower any LLM with function calling capabilities.
+# Copyright (C) 2025 FunnyCups (https://github.com/funnycups)
+
+import os
+import yaml
+from typing import List, Dict, Any, Set, Optional
+from pydantic import BaseModel, Field, field_validator
+
+
+class ServerConfig(BaseModel):
+    """Server configuration"""
+    port: int = Field(default=8000, ge=1, le=65535, description="Server port")
+    host: str = Field(default="0.0.0.0", description="Server host address")
+    timeout: int = Field(default=180, ge=1, description="Request timeout (seconds)")
+
+
+class UpstreamService(BaseModel):
+    """Upstream service configuration"""
+    name: str = Field(description="Service name")
+    base_url: str = Field(description="Service base URL")
+    api_key: str = Field(description="API key")
+    models: List[str] = Field(description="List of supported models")
+    description: str = Field(default="", description="Service description")
+    is_default: bool = Field(default=False, description="Is default service")
+    
+    @field_validator('base_url')
+    def validate_base_url(cls, v):
+        if not v.startswith(('http://', 'https://')):
+            raise ValueError('base_url must start with http:// or https://')
+        return v.rstrip('/')
+    
+    @field_validator('api_key')
+    def validate_api_key(cls, v):
+        if not v or v.strip() == "":
+            raise ValueError('api_key cannot be empty')
+        return v
+    
+    @field_validator('models')
+    def validate_models(cls, v):
+        if not v or len(v) == 0:
+            raise ValueError('models list cannot be empty')
+        for model in v:
+            if not model or model.strip() == "":
+                raise ValueError('model name cannot be empty')
+        return v
+
+
+class ClientAuthConfig(BaseModel):
+    """Client authentication configuration"""
+    allowed_keys: List[str] = Field(description="List of allowed client API keys")
+    
+    @field_validator('allowed_keys')
+    def validate_allowed_keys(cls, v):
+        if not v or len(v) == 0:
+            raise ValueError('allowed_keys cannot be empty')
+        for key in v:
+            if not key or key.strip() == "":
+                raise ValueError('API key cannot be empty')
+        return v
+
+
+class FeaturesConfig(BaseModel):
+    """Feature configuration"""
+    enable_function_calling: bool = Field(default=True, description="Enable function calling")
+    log_level: str = Field(default="INFO", description="Logging level: DEBUG, INFO, WARNING, ERROR, CRITICAL, or DISABLED")
+    convert_developer_to_system: bool = Field(default=True, description="Convert developer role to system role")
+    prompt_template: Optional[str] = Field(default=None, description="Custom prompt template for function calling")
+    key_passthrough: bool = Field(default=False, description="If true, directly forward client-provided API key to upstream instead of using configured upstream key")
+    model_passthrough: bool = Field(default=False, description="If true, forward all requests directly to the 'openai' upstream service, ignoring model-based routing")
+
+    @field_validator('log_level')
+    def validate_log_level(cls, v):
+        valid_levels = ["DEBUG", "INFO", "WARNING", "ERROR", "CRITICAL", "DISABLED"]
+        if v.upper() not in valid_levels:
+            raise ValueError(f"log_level must be one of {valid_levels}")
+        return v.upper()
+
+    @field_validator('prompt_template')
+    def validate_prompt_template(cls, v):
+        if v:
+            if "{tools_list}" not in v or "{trigger_signal}" not in v:
+                raise ValueError("prompt_template must contain {tools_list} and {trigger_signal} placeholders")
+        return v
+
+
+class AppConfig(BaseModel):
+    """Application full configuration"""
+    server: ServerConfig = Field(default_factory=ServerConfig)
+    upstream_services: List[UpstreamService] = Field(description="List of upstream services")
+    client_authentication: ClientAuthConfig = Field(description="Client authentication configuration")
+    features: FeaturesConfig = Field(default_factory=FeaturesConfig)
+    
+    @field_validator('upstream_services')
+    def validate_upstream_services(cls, v):
+        if not v or len(v) == 0:
+            raise ValueError('upstream_services cannot be empty')
+        
+        default_services = [service for service in v if service.is_default]
+        if len(default_services) == 0:
+            raise ValueError('Must have at least one default upstream service (is_default: true)')
+        if len(default_services) > 1:
+            raise ValueError('Only one upstream service can be marked as default')
+        
+        all_models = set()
+        all_aliases = set()
+        
+        for service in v:
+            for model in service.models:
+                if model in all_models:
+                    raise ValueError(f'Duplicate model entry found: {model}')
+                all_models.add(model)
+                
+                if ':' in model:
+                    parts = model.split(':', 1)
+                    if len(parts) == 2:
+                        alias, real_model = parts
+                        if not alias.strip() or not real_model.strip():
+                            raise ValueError(f"Invalid alias format in '{model}'. Both parts must not be empty.")
+                        all_aliases.add(alias)
+                    else:
+                        raise ValueError(f"Invalid model format with colon: {model}")
+
+        regular_models = {m for m in all_models if ':' not in m}
+        conflicts = all_aliases.intersection(regular_models)
+        if conflicts:
+            raise ValueError(f"Alias names {conflicts} conflict with model names.")
+                
+        return v
+
+
+class ConfigLoader:
+    """Configuration loader"""
+    
+    def __init__(self, config_path: str = "config.yaml"):
+        self.config_path = config_path
+        self._config: AppConfig = None
+    
+    def load_config(self) -> AppConfig:
+        """Load configuration file"""
+        if not os.path.exists(self.config_path):
+            raise FileNotFoundError(
+                f"Configuration file '{self.config_path}' not found. "
+                f"Please copy 'config.example.yaml' to '{self.config_path}' and modify the configuration as needed."
+            )
+        
+        try:
+            with open(self.config_path, 'r', encoding='utf-8') as f:
+                config_data = yaml.safe_load(f)
+        except yaml.YAMLError as e:
+            raise ValueError(f"Configuration file format error: {e}")
+        except Exception as e:
+            raise ValueError(f"Failed to read configuration file: {e}")
+        
+        if not config_data:
+            raise ValueError("Configuration file is empty")
+        
+        try:
+            self._config = AppConfig(**config_data)
+            return self._config
+        except Exception as e:
+            raise ValueError(f"Configuration validation failed: {e}")
+    
+    @property
+    def config(self) -> AppConfig:
+        """Get configuration object"""
+        if self._config is None:
+            self.load_config()
+        return self._config
+    
+    def get_model_to_service_mapping(self) -> tuple[Dict[str, Dict[str, Any]], Dict[str, List[str]]]:
+        """Get model to service mapping and model aliases"""
+        config = self.config
+        model_mapping = {}
+        alias_mapping = {}
+        
+        for service in config.upstream_services:
+            service_info = {
+                "name": service.name,
+                "base_url": service.base_url,
+                "api_key": service.api_key,
+                "description": service.description,
+                "is_default": service.is_default
+            }
+            
+            for model_entry in service.models:
+                model_mapping[model_entry] = service_info
+                if ':' in model_entry:
+                    parts = model_entry.split(':', 1)
+                    if len(parts) == 2:
+                        alias, _ = parts
+                        if alias not in alias_mapping:
+                            alias_mapping[alias] = []
+                        alias_mapping[alias].append(model_entry)
+        
+        return model_mapping, alias_mapping
+    
+    def get_default_service(self) -> Dict[str, Any]:
+        """Get default service configuration"""
+        config = self.config
+        for service in config.upstream_services:
+            if service.is_default:
+                return {
+                    "name": service.name,
+                    "base_url": service.base_url,
+                    "api_key": service.api_key,
+                    "description": service.description,
+                    "is_default": service.is_default
+                }
+        raise ValueError("No default service configured")
+    
+    def get_allowed_client_keys(self) -> Set[str]:
+        """Get set of allowed client keys"""
+        return set(self.config.client_authentication.allowed_keys)
+    
+    def get_log_level(self) -> str:
+        """Get configured log level"""
+        return self.config.features.log_level
+    
+    def get_features_config(self) -> Dict[str, Any]:
+        """Get feature configuration"""
+        return {
+            "function_calling": self.config.features.enable_function_calling,
+            "log_level": self.config.features.log_level,
+            "convert_developer_to_system": self.config.features.convert_developer_to_system
+        }
+
+
+config_loader = ConfigLoader()

docker-compose.yml

@@ -0,0 +1,14 @@
+version: '3.8'
+
+services:
+  toolify:
+    image: toolify
+    build:
+      context: .
+      dockerfile: Dockerfile
+    container_name: toolify
+    ports:
+      - "8000:8000"
+    volumes:
+      - ./config.yaml:/app/config.yaml
+    restart: unless-stopped

docker-entrypoint.sh

@@ -0,0 +1,18 @@
+#!/usr/bin/env bash
+set -e
+
+# 如果缺少配置文件且允许自动生成，则触发脚本生成
+PORT="${PORT:-8000}"
+CONFIG_PATH="${CONFIG_PATH:-/app/config.yaml}"
+GENERATE_CONFIG="${GENERATE_CONFIG:-0}"
+
+if [ ! -f "$CONFIG_PATH" ]; then
+  if [ "$GENERATE_CONFIG" != "0" ]; then
+    echo "正在基于环境变量生成配置文件..."
+    python -m scripts.generate_config
+  else
+    echo "警告: 未找到配置文件 $CONFIG_PATH，且未启用自动生成。" >&2
+  fi
+fi
+
+exec uvicorn main:app --host 0.0.0.0 --port "$PORT"

main.py

@@ -0,0 +1,1360 @@
+# SPDX-License-Identifier: GPL-3.0-or-later
+#
+# Toolify: Empower any LLM with function calling capabilities.
+# Copyright (C) 2025 FunnyCups (https://github.com/funnycups)
+
+import os
+import re
+import json
+import uuid
+import httpx
+import secrets
+import string
+import traceback
+import time
+import random
+import threading
+import logging
+from typing import List, Dict, Any, Optional, Literal, Union
+from collections import OrderedDict
+
+from fastapi import FastAPI, Request, Header, HTTPException, Depends
+from fastapi.responses import JSONResponse, StreamingResponse
+from pydantic import BaseModel, ValidationError
+
+from config_loader import config_loader
+
+logger = logging.getLogger(__name__)
+
+def generate_random_trigger_signal() -> str:
+    """Generate a random, self-closing trigger signal like <Function_AB1c_Start/>."""
+    chars = string.ascii_letters + string.digits
+    random_str = ''.join(secrets.choice(chars) for _ in range(4))
+    return f"<Function_{random_str}_Start/>"
+
+try:
+    app_config = config_loader.load_config()
+    
+    log_level_str = app_config.features.log_level
+    if log_level_str == "DISABLED":
+        log_level = logging.CRITICAL + 1
+    else:
+        log_level = getattr(logging, log_level_str, logging.INFO)
+    
+    logging.basicConfig(
+        level=log_level,
+        format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
+        datefmt='%Y-%m-%d %H:%M:%S'
+    )
+    
+    logger.info(f"✅ Configuration loaded successfully: {config_loader.config_path}")
+    logger.info(f"📊 Configured {len(app_config.upstream_services)} upstream services")
+    logger.info(f"🔑 Configured {len(app_config.client_authentication.allowed_keys)} client keys")
+    
+    MODEL_TO_SERVICE_MAPPING, ALIAS_MAPPING = config_loader.get_model_to_service_mapping()
+    DEFAULT_SERVICE = config_loader.get_default_service()
+    ALLOWED_CLIENT_KEYS = config_loader.get_allowed_client_keys()
+    GLOBAL_TRIGGER_SIGNAL = generate_random_trigger_signal()
+    
+    logger.info(f"🎯 Configured {len(MODEL_TO_SERVICE_MAPPING)} model mappings")
+    if ALIAS_MAPPING:
+        logger.info(f"🔄 Configured {len(ALIAS_MAPPING)} model aliases: {list(ALIAS_MAPPING.keys())}")
+    logger.info(f"🔄 Default service: {DEFAULT_SERVICE['name']}")
+    
+except Exception as e:
+    logger.error(f"❌ Configuration loading failed: {type(e).__name__}")
+    logger.error(f"❌ Error details: {str(e)}")
+    logger.error("💡 Please ensure config.yaml file exists and is properly formatted")
+    exit(1)
+class ToolCallMappingManager:
+    """
+    Tool call mapping manager with TTL (Time To Live) and size limit
+    
+    Features:
+    1. Automatic expiration cleanup - entries are automatically deleted after specified time
+    2. Size limit - prevents unlimited memory growth
+    3. LRU eviction - removes least recently used entries when size limit is reached
+    4. Thread safe - supports concurrent access
+    5. Periodic cleanup - background thread regularly cleans up expired entries
+    """
+    
+    def __init__(self, max_size: int = 1000, ttl_seconds: int = 3600, cleanup_interval: int = 300):
+        """
+        Initialize mapping manager
+        
+        Args:
+            max_size: Maximum number of stored entries
+            ttl_seconds: Entry time to live (seconds)
+            cleanup_interval: Cleanup interval (seconds)
+        """
+        self.max_size = max_size
+        self.ttl_seconds = ttl_seconds
+        self.cleanup_interval = cleanup_interval
+        
+        self._data: OrderedDict[str, Dict[str, Any]] = OrderedDict()
+        self._timestamps: Dict[str, float] = {}
+        self._lock = threading.RLock()
+        
+        self._cleanup_thread = threading.Thread(target=self._periodic_cleanup, daemon=True)
+        self._cleanup_thread.start()
+        
+        logger.debug(f"🔧 [INIT] Tool call mapping manager started - Max entries: {max_size}, TTL: {ttl_seconds}s, Cleanup interval: {cleanup_interval}s")
+    
+    def store(self, tool_call_id: str, name: str, args: dict, description: str = "") -> None:
+        """Store tool call mapping"""
+        with self._lock:
+            current_time = time.time()
+            
+            if tool_call_id in self._data:
+                del self._data[tool_call_id]
+                del self._timestamps[tool_call_id]
+            
+            while len(self._data) >= self.max_size:
+                oldest_key = next(iter(self._data))
+                del self._data[oldest_key]
+                del self._timestamps[oldest_key]
+                logger.debug(f"🔧 [CLEANUP] Removed oldest entry due to size limit: {oldest_key}")
+            
+            self._data[tool_call_id] = {
+                "name": name,
+                "args": args,
+                "description": description,
+                "created_at": current_time
+            }
+            self._timestamps[tool_call_id] = current_time
+            
+            logger.debug(f"🔧 Stored tool call mapping: {tool_call_id} -> {name}")
+            logger.debug(f"🔧 Current mapping table size: {len(self._data)}")
+    
+    def get(self, tool_call_id: str) -> Optional[Dict[str, Any]]:
+        """Get tool call mapping (updates LRU order)"""
+        with self._lock:
+            current_time = time.time()
+            
+            if tool_call_id not in self._data:
+                logger.debug(f"🔧 Tool call mapping not found: {tool_call_id}")
+                logger.debug(f"🔧 All IDs in current mapping table: {list(self._data.keys())}")
+                return None
+            
+            if current_time - self._timestamps[tool_call_id] > self.ttl_seconds:
+                logger.debug(f"🔧 Tool call mapping expired: {tool_call_id}")
+                del self._data[tool_call_id]
+                del self._timestamps[tool_call_id]
+                return None
+            
+            result = self._data[tool_call_id]
+            self._data.move_to_end(tool_call_id)
+            
+            logger.debug(f"🔧 Found tool call mapping: {tool_call_id} -> {result['name']}")
+            return result
+    
+    def cleanup_expired(self) -> int:
+        """Clean up expired entries, return the number of cleaned entries"""
+        with self._lock:
+            current_time = time.time()
+            expired_keys = []
+            
+            for key, timestamp in self._timestamps.items():
+                if current_time - timestamp > self.ttl_seconds:
+                    expired_keys.append(key)
+            
+            for key in expired_keys:
+                del self._data[key]
+                del self._timestamps[key]
+            
+            if expired_keys:
+                logger.debug(f"🔧 [CLEANUP] Cleaned up {len(expired_keys)} expired entries")
+            
+            return len(expired_keys)
+    
+    def get_stats(self) -> Dict[str, Any]:
+        """Get statistics"""
+        with self._lock:
+            current_time = time.time()
+            expired_count = sum(1 for ts in self._timestamps.values()
+                              if current_time - ts > self.ttl_seconds)
+            
+            return {
+                "total_entries": len(self._data),
+                "expired_entries": expired_count,
+                "active_entries": len(self._data) - expired_count,
+                "max_size": self.max_size,
+                "ttl_seconds": self.ttl_seconds,
+                "memory_usage_ratio": len(self._data) / self.max_size
+            }
+    
+    def _periodic_cleanup(self) -> None:
+        """Background periodic cleanup thread"""
+        while True:
+            try:
+                time.sleep(self.cleanup_interval)
+                cleaned = self.cleanup_expired()
+                
+                stats = self.get_stats()
+                if stats["total_entries"] > 0:
+                    logger.debug(f"🔧 [STATS] Mapping table status: Total={stats['total_entries']}, "
+                               f"Active={stats['active_entries']}, Memory usage={stats['memory_usage_ratio']:.1%}")
+                
+            except Exception as e:
+                logger.error(f"❌ Background cleanup thread exception: {e}")
+
+TOOL_CALL_MAPPING_MANAGER = ToolCallMappingManager(
+    max_size=1000,
+    ttl_seconds=3600,
+    cleanup_interval=300
+)
+
+def store_tool_call_mapping(tool_call_id: str, name: str, args: dict, description: str = ""):
+    """Store mapping between tool call ID and call content"""
+    TOOL_CALL_MAPPING_MANAGER.store(tool_call_id, name, args, description)
+
+def get_tool_call_mapping(tool_call_id: str) -> Optional[Dict[str, Any]]:
+    """Get call content corresponding to tool call ID"""
+    return TOOL_CALL_MAPPING_MANAGER.get(tool_call_id)
+
+def format_tool_result_for_ai(tool_call_id: str, result_content: str) -> str:
+    """Format tool call results for AI understanding with English prompts and XML structure"""
+    logger.debug(f"🔧 Formatting tool call result: tool_call_id={tool_call_id}")
+    tool_info = get_tool_call_mapping(tool_call_id)
+    if not tool_info:
+        logger.debug(f"🔧 Tool call mapping not found, using default format")
+        return f"Tool execution result:\n<tool_result>\n{result_content}\n</tool_result>"
+    
+    formatted_text = f"""Tool execution result:
+- Tool name: {tool_info['name']}
+- Execution result:
+<tool_result>
+{result_content}
+</tool_result>"""
+    
+    logger.debug(f"🔧 Formatting completed, tool name: {tool_info['name']}")
+    return formatted_text
+
+def format_assistant_tool_calls_for_ai(tool_calls: List[Dict[str, Any]], trigger_signal: str) -> str:
+    """Format assistant tool calls into AI-readable string format."""
+    logger.debug(f"🔧 Formatting assistant tool calls. Count: {len(tool_calls)}")
+    
+    xml_calls_parts = []
+    for tool_call in tool_calls:
+        function_info = tool_call.get("function", {})
+        name = function_info.get("name", "")
+        arguments_json = function_info.get("arguments", "{}")
+        
+        try:
+            # First, try to load as JSON. If it's a string that's a valid JSON, we parse it.
+            args_dict = json.loads(arguments_json)
+        except (json.JSONDecodeError, TypeError):
+            # If it's not a valid JSON string, treat it as a simple string.
+            args_dict = {"raw_arguments": arguments_json}
+
+        args_parts = []
+        for key, value in args_dict.items():
+            # Dump the value back to a JSON string for consistent representation inside XML.
+            json_value = json.dumps(value, ensure_ascii=False)
+            args_parts.append(f"<{key}>{json_value}</{key}>")
+        
+        args_content = "\n".join(args_parts)
+        
+        xml_call = f"<function_call>\n<tool>{name}</tool>\n<args>\n{args_content}\n</args>\n</function_call>"
+        xml_calls_parts.append(xml_call)
+
+    all_calls = "\n".join(xml_calls_parts)
+    final_str = f"{trigger_signal}\n<function_calls>\n{all_calls}\n</function_calls>"
+    
+    logger.debug("🔧 Assistant tool calls formatted successfully.")
+    return final_str
+
+def get_function_call_prompt_template(trigger_signal: str) -> str:
+    """
+    Generate prompt template based on dynamic trigger signal
+    """
+    custom_template = app_config.features.prompt_template
+    if custom_template:
+        logger.info("🔧 Using custom prompt template from configuration")
+        return custom_template.format(
+            trigger_signal=trigger_signal,
+            tools_list="{tools_list}"
+        )
+    
+    return f"""
+You have access to the following available tools to help solve problems:
+
+{{tools_list}}
+
+**IMPORTANT CONTEXT NOTES:**
+1. You can call MULTIPLE tools in a single response if needed.
+2. The conversation context may already contain tool execution results from previous function calls. Review the conversation history carefully to avoid unnecessary duplicate tool calls.
+3. When tool execution results are present in the context, they will be formatted with XML tags like <tool_result>...</tool_result> for easy identification.
+4. This is the ONLY format you can use for tool calls, and any deviation will result in failure.
+
+When you need to use tools, you **MUST** strictly follow this format. Do NOT include any extra text, explanations, or dialogue on the first and second lines of the tool call syntax:
+
+1. When starting tool calls, begin on a new line with exactly:
+{trigger_signal}
+No leading or trailing spaces, output exactly as shown above. The trigger signal MUST be on its own line and appear only once.
+
+2. Starting from the second line, **immediately** follow with the complete <function_calls> XML block.
+
+3. For multiple tool calls, include multiple <function_call> blocks within the same <function_calls> wrapper.
+
+4. Do not add any text or explanation after the closing </function_calls> tag.
+
+STRICT ARGUMENT KEY RULES:
+- You MUST use parameter keys EXACTLY as defined (case- and punctuation-sensitive). Do NOT rename, add, or remove characters.
+- If a key starts with a hyphen (e.g., -i, -C), you MUST keep the hyphen in the tag name. Example: <-i>true</-i>, <-C>2</-C>.
+- Never convert "-i" to "i" or "-C" to "C". Do not pluralize, translate, or alias parameter keys.
+- The <tool> tag must contain the exact name of a tool from the list. Any other tool name is invalid.
+- The <args> must contain all required arguments for that tool.
+
+CORRECT Example (multiple tool calls, including hyphenated keys):
+...response content (optional)...
+{trigger_signal}
+<function_calls>
+    <function_call>
+        <tool>Grep</tool>
+        <args>
+            <-i>true</-i>
+            <-C>2</-C>
+            <path>.</path>
+        </args>
+    </function_call>
+    <function_call>
+        <tool>search</tool>
+        <args>
+            <keywords>["Python Document", "how to use python"]</keywords>
+        </args>
+    </function_call>
+  </function_calls>
+
+INCORRECT Example (extra text + wrong key names — DO NOT DO THIS):
+...response content (optional)...
+{trigger_signal}
+I will call the tools for you.
+<function_calls>
+    <function_call>
+        <tool>Grep</tool>
+        <args>
+            <i>true</i>
+            <C>2</C>
+            <path>.</path>
+        </args>
+    </function_call>
+</function_calls>
+
+Now please be ready to strictly follow the above specifications.
+"""
+
+class ToolFunction(BaseModel):
+    name: str
+    description: Optional[str] = None
+    parameters: Dict[str, Any]
+
+class Tool(BaseModel):
+    type: Literal["function"]
+    function: ToolFunction
+
+class Message(BaseModel):
+    role: str
+    content: Optional[str] = None
+    tool_calls: Optional[List[Dict[str, Any]]] = None
+    tool_call_id: Optional[str] = None
+    name: Optional[str] = None
+    
+    class Config:
+        extra = "allow"
+
+class ToolChoice(BaseModel):
+    type: Literal["function"]
+    function: Dict[str, str]
+
+class ChatCompletionRequest(BaseModel):
+    model: str
+    messages: List[Dict[str, Any]]
+    tools: Optional[List[Tool]] = None
+    tool_choice: Optional[Union[str, ToolChoice]] = None
+    stream: Optional[bool] = False
+    stream_options: Optional[Dict[str, Any]] = None
+    temperature: Optional[float] = None
+    max_tokens: Optional[int] = None
+    top_p: Optional[float] = None
+    frequency_penalty: Optional[float] = None
+    presence_penalty: Optional[float] = None
+    n: Optional[int] = None
+    stop: Optional[Union[str, List[str]]] = None
+    
+    class Config:
+        extra = "allow"
+
+
+def generate_function_prompt(tools: List[Tool], trigger_signal: str) -> tuple[str, str]:
+    """
+    Generate injected system prompt based on tools definition in client request.
+    Returns: (prompt_content, trigger_signal)
+    """
+    tools_list_str = []
+    for i, tool in enumerate(tools):
+        func = tool.function
+        name = func.name
+        description = func.description or ""
+
+        # Robustly read JSON Schema fields
+        schema: Dict[str, Any] = func.parameters or {}
+        props: Dict[str, Any] = schema.get("properties", {}) or {}
+        required_list: List[str] = schema.get("required", []) or []
+
+        # Brief summary line: name (type)
+        params_summary = ", ".join([
+            f"{p_name} ({(p_info or {}).get('type', 'any')})" for p_name, p_info in props.items()
+        ]) or "None"
+
+        # Build detailed parameter spec for prompt injection (default enabled)
+        detail_lines: List[str] = []
+        for p_name, p_info in props.items():
+            p_info = p_info or {}
+            p_type = p_info.get("type", "any")
+            is_required = "Yes" if p_name in required_list else "No"
+            p_desc = p_info.get("description")
+            enum_vals = p_info.get("enum")
+            default_val = p_info.get("default")
+            examples_val = p_info.get("examples") or p_info.get("example")
+
+            # Common constraints and hints
+            constraints: Dict[str, Any] = {}
+            for key in [
+                "minimum", "maximum", "exclusiveMinimum", "exclusiveMaximum",
+                "minLength", "maxLength", "pattern", "format",
+                "minItems", "maxItems", "uniqueItems"
+            ]:
+                if key in p_info:
+                    constraints[key] = p_info.get(key)
+
+            # Array item type hint
+            if p_type == "array":
+                items = p_info.get("items") or {}
+                if isinstance(items, dict):
+                    itype = items.get("type")
+                    if itype:
+                        constraints["items.type"] = itype
+
+            # Compose pretty lines
+            detail_lines.append(f"- {p_name}:")
+            detail_lines.append(f"  - type: {p_type}")
+            detail_lines.append(f"  - required: {is_required}")
+            if p_desc:
+                detail_lines.append(f"  - description: {p_desc}")
+            if enum_vals is not None:
+                try:
+                    detail_lines.append(f"  - enum: {json.dumps(enum_vals, ensure_ascii=False)}")
+                except Exception:
+                    detail_lines.append(f"  - enum: {enum_vals}")
+            if default_val is not None:
+                try:
+                    detail_lines.append(f"  - default: {json.dumps(default_val, ensure_ascii=False)}")
+                except Exception:
+                    detail_lines.append(f"  - default: {default_val}")
+            if examples_val is not None:
+                try:
+                    detail_lines.append(f"  - examples: {json.dumps(examples_val, ensure_ascii=False)}")
+                except Exception:
+                    detail_lines.append(f"  - examples: {examples_val}")
+            if constraints:
+                try:
+                    detail_lines.append(f"  - constraints: {json.dumps(constraints, ensure_ascii=False)}")
+                except Exception:
+                    detail_lines.append(f"  - constraints: {constraints}")
+
+        detail_block = "\n".join(detail_lines) if detail_lines else "(no parameter details)"
+
+        desc_block = f"```\n{description}\n```" if description else "None"
+
+        tools_list_str.append(
+            f"{i + 1}. <tool name=\"{name}\">\n"
+            f"   Description:\n{desc_block}\n"
+            f"   Parameters summary: {params_summary}\n"
+            f"   Required parameters: {', '.join(required_list) if required_list else 'None'}\n"
+            f"   Parameter details:\n{detail_block}"
+        )
+    
+    prompt_template = get_function_call_prompt_template(trigger_signal)
+    prompt_content = prompt_template.replace("{tools_list}", "\n\n".join(tools_list_str))
+    
+    return prompt_content, trigger_signal
+
+def remove_think_blocks(text: str) -> str:
+    """
+    Temporarily remove all <think>...</think> blocks for XML parsing
+    Supports nested think tags
+    Note: This function is only used for temporary parsing and does not affect the original content returned to the user
+    """
+    while '<think>' in text and '</think>' in text:
+        start_pos = text.find('<think>')
+        if start_pos == -1:
+            break
+        
+        pos = start_pos + 7
+        depth = 1
+        
+        while pos < len(text) and depth > 0:
+            if text[pos:pos+7] == '<think>':
+                depth += 1
+                pos += 7
+            elif text[pos:pos+8] == '</think>':
+                depth -= 1
+                pos += 8
+            else:
+                pos += 1
+        
+        if depth == 0:
+            text = text[:start_pos] + text[pos:]
+        else:
+            break
+    
+    return text
+
+class StreamingFunctionCallDetector:
+    """Enhanced streaming function call detector, supports dynamic trigger signals, avoids misjudgment within <think> tags
+    
+    Core features:
+    1. Avoid triggering tool call detection within <think> blocks
+    2. Normally output <think> block content to the user
+    3. Supports nested think tags
+    """
+    
+    def __init__(self, trigger_signal: str):
+        self.trigger_signal = trigger_signal
+        self.reset()
+    
+    def reset(self):
+        self.content_buffer = ""
+        self.state = "detecting"  # detecting, tool_parsing
+        self.in_think_block = False
+        self.think_depth = 0
+        self.signal = self.trigger_signal
+        self.signal_len = len(self.signal)
+    
+    def process_chunk(self, delta_content: str) -> tuple[bool, str]:
+        """
+        Process streaming content chunk
+        Returns: (is_tool_call_detected, content_to_yield)
+        """
+        if not delta_content:
+            return False, ""
+        
+        self.content_buffer += delta_content
+        content_to_yield = ""
+        
+        if self.state == "tool_parsing":
+            return False, ""
+        
+        if delta_content:
+            logger.debug(f"🔧 Processing chunk: {repr(delta_content[:50])}{'...' if len(delta_content) > 50 else ''}, buffer length: {len(self.content_buffer)}, think state: {self.in_think_block}")
+        
+        i = 0
+        while i < len(self.content_buffer):
+            skip_chars = self._update_think_state(i)
+            if skip_chars > 0:
+                for j in range(skip_chars):
+                    if i + j < len(self.content_buffer):
+                        content_to_yield += self.content_buffer[i + j]
+                i += skip_chars
+                continue
+            
+            if not self.in_think_block and self._can_detect_signal_at(i):
+                if self.content_buffer[i:i+self.signal_len] == self.signal:
+                    logger.debug(f"🔧 Improved detector: detected trigger signal in non-think block! Signal: {self.signal[:20]}...")
+                    logger.debug(f"🔧 Trigger signal position: {i}, think state: {self.in_think_block}, think depth: {self.think_depth}")
+                    self.state = "tool_parsing"
+                    self.content_buffer = self.content_buffer[i:]
+                    return True, content_to_yield
+            
+            remaining_len = len(self.content_buffer) - i
+            if remaining_len < self.signal_len or remaining_len < 8:
+                break
+            
+            content_to_yield += self.content_buffer[i]
+            i += 1
+        
+        self.content_buffer = self.content_buffer[i:]
+        return False, content_to_yield
+    
+    def _update_think_state(self, pos: int):
+        """Update think tag state, supports nesting"""
+        remaining = self.content_buffer[pos:]
+        
+        if remaining.startswith('<think>'):
+            self.think_depth += 1
+            self.in_think_block = True
+            logger.debug(f"🔧 Entering think block, depth: {self.think_depth}")
+            return 7
+        
+        elif remaining.startswith('</think>'):
+            self.think_depth = max(0, self.think_depth - 1)
+            self.in_think_block = self.think_depth > 0
+            logger.debug(f"🔧 Exiting think block, depth: {self.think_depth}")
+            return 8
+        
+        return 0
+    
+    def _can_detect_signal_at(self, pos: int) -> bool:
+        """Check if signal can be detected at the specified position"""
+        return (pos + self.signal_len <= len(self.content_buffer) and 
+                not self.in_think_block)
+    
+    def finalize(self) -> Optional[List[Dict[str, Any]]]:
+        """Final processing when stream ends"""
+        if self.state == "tool_parsing":
+            return parse_function_calls_xml(self.content_buffer, self.trigger_signal)
+        return None
+
+def parse_function_calls_xml(xml_string: str, trigger_signal: str) -> Optional[List[Dict[str, Any]]]:
+    """
+    Enhanced XML parsing function, supports dynamic trigger signals
+    1. Retain <think>...</think> blocks (they should be returned normally to the user)
+    2. Temporarily remove think blocks only when parsing function_calls to prevent think content from interfering with XML parsing
+    3. Find the last occurrence of the trigger signal
+    4. Start parsing function_calls from the last trigger signal
+    """
+    logger.debug(f"🔧 Improved parser starting processing, input length: {len(xml_string) if xml_string else 0}")
+    logger.debug(f"🔧 Using trigger signal: {trigger_signal[:20]}...")
+    
+    if not xml_string or trigger_signal not in xml_string:
+        logger.debug(f"🔧 Input is empty or doesn't contain trigger signal")
+        return None
+    
+    cleaned_content = remove_think_blocks(xml_string)
+    logger.debug(f"🔧 Content length after temporarily removing think blocks: {len(cleaned_content)}")
+    
+    signal_positions = []
+    start_pos = 0
+    while True:
+        pos = cleaned_content.find(trigger_signal, start_pos)
+        if pos == -1:
+            break
+        signal_positions.append(pos)
+        start_pos = pos + 1
+    
+    if not signal_positions:
+        logger.debug(f"🔧 No trigger signal found in cleaned content")
+        return None
+    
+    logger.debug(f"🔧 Found {len(signal_positions)} trigger signal positions: {signal_positions}")
+    
+    last_signal_pos = signal_positions[-1]
+    content_after_signal = cleaned_content[last_signal_pos:]
+    logger.debug(f"🔧 Content starting from last trigger signal: {repr(content_after_signal[:100])}")
+    
+    calls_content_match = re.search(r"<function_calls>([\s\S]*?)</function_calls>", content_after_signal)
+    if not calls_content_match:
+        logger.debug(f"🔧 No function_calls tag found")
+        return None
+    
+    calls_content = calls_content_match.group(1)
+    logger.debug(f"🔧 function_calls content: {repr(calls_content)}")
+    
+    results = []
+    call_blocks = re.findall(r"<function_call>([\s\S]*?)</function_call>", calls_content)
+    logger.debug(f"🔧 Found {len(call_blocks)} function_call blocks")
+    
+    for i, block in enumerate(call_blocks):
+        logger.debug(f"🔧 Processing function_call #{i+1}: {repr(block)}")
+        
+        tool_match = re.search(r"<tool>(.*?)</tool>", block)
+        if not tool_match:
+            logger.debug(f"🔧 No tool tag found in block #{i+1}")
+            continue
+        
+        name = tool_match.group(1).strip()
+        args = {}
+        
+        args_block_match = re.search(r"<args>([\s\S]*?)</args>", block)
+        if args_block_match:
+            args_content = args_block_match.group(1)
+            # Support arg tag names containing hyphens (e.g., -i, -A); match any non-space, non-'>' and non-'/' chars
+            arg_matches = re.findall(r"<([^\s>/]+)>([\s\S]*?)</\1>", args_content)
+
+            def _coerce_value(v: str):
+                try:
+                    return json.loads(v)
+                except Exception:
+                    pass
+                return v
+
+            for k, v in arg_matches:
+                args[k] = _coerce_value(v)
+        
+        result = {"name": name, "args": args}
+        results.append(result)
+        logger.debug(f"🔧 Added tool call: {result}")
+    
+    logger.debug(f"🔧 Final parsing result: {results}")
+    return results if results else None
+
+def find_upstream(model_name: str) -> tuple[Dict[str, Any], str]:
+    """Find upstream configuration by model name, handling aliases and passthrough mode."""
+    
+    # Handle model passthrough mode
+    if app_config.features.model_passthrough:
+        logger.info("🔄 Model passthrough mode is active. Forwarding to 'openai' service.")
+        openai_service = None
+        for service in app_config.upstream_services:
+            if service.name == "openai":
+                openai_service = service.model_dump()
+                break
+        
+        if openai_service:
+            if not openai_service.get("api_key"):
+                 raise HTTPException(status_code=500, detail="Configuration error: API key not found for the 'openai' service in model passthrough mode.")
+            # In passthrough mode, the model name from the request is used directly.
+            return openai_service, model_name
+        else:
+            raise HTTPException(status_code=500, detail="Configuration error: 'model_passthrough' is enabled, but no upstream service named 'openai' was found.")
+
+    # Default routing logic
+    chosen_model_entry = model_name
+    
+    if model_name in ALIAS_MAPPING:
+        chosen_model_entry = random.choice(ALIAS_MAPPING[model_name])
+        logger.info(f"🔄 Model alias '{model_name}' detected. Randomly selected '{chosen_model_entry}' for this request.")
+
+    service = MODEL_TO_SERVICE_MAPPING.get(chosen_model_entry)
+    
+    if service:
+        if not service.get("api_key"):
+            raise HTTPException(status_code=500, detail=f"Model configuration error: API key not found for service '{service.get('name')}'.")
+    else:
+        logger.warning(f"⚠️  Model '{model_name}' not found in configuration, using default service")
+        service = DEFAULT_SERVICE
+        if not service.get("api_key"):
+            raise HTTPException(status_code=500, detail="Service configuration error: Default API key not found.")
+
+    actual_model_name = chosen_model_entry
+    if ':' in chosen_model_entry:
+         parts = chosen_model_entry.split(':', 1)
+         if len(parts) == 2:
+             _, actual_model_name = parts
+            
+    return service, actual_model_name
+
+app = FastAPI()
+http_client = httpx.AsyncClient()
+
+@app.middleware("http")
+async def debug_middleware(request: Request, call_next):
+    """Middleware for debugging validation errors, does not log conversation content."""
+    response = await call_next(request)
+    
+    if response.status_code == 422:
+        logger.debug(f"🔍 Validation error detected for {request.method} {request.url.path}")
+        logger.debug(f"🔍 Response status code: 422 (Pydantic validation failure)")
+    
+    return response
+
+@app.exception_handler(ValidationError)
+async def validation_exception_handler(request: Request, exc: ValidationError):
+    """Handle Pydantic validation errors with detailed error information"""
+    logger.error(f"❌ Pydantic validation error: {exc}")
+    logger.error(f"❌ Request URL: {request.url}")
+    logger.error(f"❌ Error details: {exc.errors()}")
+    
+    for error in exc.errors():
+        logger.error(f"❌ Validation error location: {error.get('loc')}")
+        logger.error(f"❌ Validation error message: {error.get('msg')}")
+        logger.error(f"❌ Validation error type: {error.get('type')}")
+    
+    return JSONResponse(
+        status_code=422,
+        content={
+            "error": {
+                "message": "Invalid request format",
+                "type": "invalid_request_error",
+                "code": "invalid_request"
+            }
+        }
+    )
+
+@app.exception_handler(Exception)
+async def general_exception_handler(request: Request, exc: Exception):
+    """Handle all uncaught exceptions"""
+    logger.error(f"❌ Unhandled exception: {exc}")
+    logger.error(f"❌ Request URL: {request.url}")
+    logger.error(f"❌ Exception type: {type(exc).__name__}")
+    logger.error(f"❌ Error stack: {traceback.format_exc()}")
+    
+    return JSONResponse(
+        status_code=500,
+        content={
+            "error": {
+                "message": "Internal server error",
+                "type": "server_error",
+                "code": "internal_error"
+            }
+        }
+    )
+
+async def verify_api_key(authorization: str = Header(...)):
+    """Dependency: verify client API key"""
+    client_key = authorization.replace("Bearer ", "")
+    if app_config.features.key_passthrough:
+        # In passthrough mode, skip allowed_keys check
+        return client_key
+    if client_key not in ALLOWED_CLIENT_KEYS:
+        raise HTTPException(status_code=401, detail="Unauthorized")
+    return client_key
+
+def preprocess_messages(messages: List[Dict[str, Any]]) -> List[Dict[str, Any]]:
+    """Preprocess messages, convert tool-type messages to AI-understandable format, return dictionary list to avoid Pydantic validation issues"""
+    processed_messages = []
+    
+    for message in messages:
+        if isinstance(message, dict):
+            if message.get("role") == "tool":
+                tool_call_id = message.get("tool_call_id")
+                content = message.get("content")
+                
+                if tool_call_id and content:
+                    formatted_content = format_tool_result_for_ai(tool_call_id, content)
+                    processed_message = {
+                        "role": "user",
+                        "content": formatted_content
+                    }
+                    processed_messages.append(processed_message)
+                    logger.debug(f"🔧 Converted tool message to user message: tool_call_id={tool_call_id}")
+                else:
+                    logger.debug(f"🔧 Skipped invalid tool message: tool_call_id={tool_call_id}, content={bool(content)}")
+            elif message.get("role") == "assistant" and "tool_calls" in message and message["tool_calls"]:
+                tool_calls = message.get("tool_calls", [])
+                formatted_tool_calls_str = format_assistant_tool_calls_for_ai(tool_calls, GLOBAL_TRIGGER_SIGNAL)
+                
+                # Combine with original content if it exists
+                original_content = message.get("content") or ""
+                final_content = f"{original_content}\n{formatted_tool_calls_str}".strip()
+
+                processed_message = {
+                    "role": "assistant",
+                    "content": final_content
+                }
+                # Copy other potential keys from the original message, except tool_calls
+                for key, value in message.items():
+                    if key not in ["role", "content", "tool_calls"]:
+                        processed_message[key] = value
+
+                processed_messages.append(processed_message)
+                logger.debug(f"🔧 Converted assistant tool_calls to content.")
+
+            elif message.get("role") == "developer":
+                if app_config.features.convert_developer_to_system:
+                    processed_message = message.copy()
+                    processed_message["role"] = "system"
+                    processed_messages.append(processed_message)
+                    logger.debug(f"🔧 Converted developer message to system message for better upstream compatibility")
+                else:
+                    processed_messages.append(message)
+                    logger.debug(f"🔧 Keeping developer role unchanged (based on configuration)")
+            else:
+                processed_messages.append(message)
+        else:
+            processed_messages.append(message)
+    
+    return processed_messages
+
+@app.post("/v1/chat/completions")
+async def chat_completions(
+    request: Request,
+    body: ChatCompletionRequest,
+    _api_key: str = Depends(verify_api_key)
+):
+    """Main chat completion endpoint, proxy and inject function calling capabilities."""
+    try:
+        logger.debug(f"🔧 Received request, model: {body.model}")
+        logger.debug(f"🔧 Number of messages: {len(body.messages)}")
+        logger.debug(f"🔧 Number of tools: {len(body.tools) if body.tools else 0}")
+        logger.debug(f"🔧 Streaming: {body.stream}")
+        
+        upstream, actual_model = find_upstream(body.model)
+        upstream_url = f"{upstream['base_url']}/chat/completions"
+        
+        logger.debug(f"🔧 Starting message preprocessing, original message count: {len(body.messages)}")
+        processed_messages = preprocess_messages(body.messages)
+        logger.debug(f"🔧 Preprocessing completed, processed message count: {len(processed_messages)}")
+        
+        if not validate_message_structure(processed_messages):
+            logger.error(f"❌ Message structure validation failed, but continuing processing")
+        
+        request_body_dict = body.model_dump(exclude_unset=True)
+        request_body_dict["model"] = actual_model
+        request_body_dict["messages"] = processed_messages
+        is_fc_enabled = app_config.features.enable_function_calling
+        has_tools_in_request = bool(body.tools)
+        has_function_call = is_fc_enabled and has_tools_in_request
+        
+        logger.debug(f"🔧 Request body constructed, message count: {len(processed_messages)}")
+        
+    except Exception as e:
+        logger.error(f"❌ Request preprocessing failed: {str(e)}")
+        logger.error(f"❌ Error type: {type(e).__name__}")
+        if hasattr(app_config, 'debug') and app_config.debug:
+            logger.error(f"❌ Error stack: {traceback.format_exc()}")
+        
+        return JSONResponse(
+            status_code=422,
+            content={
+                "error": {
+                    "message": "Invalid request format",
+                    "type": "invalid_request_error",
+                    "code": "invalid_request"
+                }
+            }
+        )
+
+    if has_function_call:
+        logger.debug(f"🔧 Using global trigger signal for this request: {GLOBAL_TRIGGER_SIGNAL}")
+        
+        function_prompt, _ = generate_function_prompt(body.tools, GLOBAL_TRIGGER_SIGNAL)
+        
+        tool_choice_prompt = safe_process_tool_choice(body.tool_choice)
+        if tool_choice_prompt:
+            function_prompt += tool_choice_prompt
+
+        system_message = {"role": "system", "content": function_prompt}
+        request_body_dict["messages"].insert(0, system_message)
+        
+        if "tools" in request_body_dict:
+            del request_body_dict["tools"]
+        if "tool_choice" in request_body_dict:
+            del request_body_dict["tool_choice"]
+
+    elif has_tools_in_request and not is_fc_enabled:
+        logger.info(f"🔧 Function calling is disabled by configuration, ignoring 'tools' and 'tool_choice' in request.")
+        if "tools" in request_body_dict:
+            del request_body_dict["tools"]
+        if "tool_choice" in request_body_dict:
+            del request_body_dict["tool_choice"]
+
+    headers = {
+        "Content-Type": "application/json",
+        "Authorization": f"Bearer {_api_key}" if app_config.features.key_passthrough else f"Bearer {upstream['api_key']}",
+        "Accept": "application/json" if not body.stream else "text/event-stream"
+    }
+
+    logger.info(f"📝 Forwarding request to upstream: {upstream['name']}")
+    logger.info(f"📝 Model: {request_body_dict.get('model', 'unknown')}, Messages: {len(request_body_dict.get('messages', []))}")
+
+    if not body.stream:
+        try:
+            logger.debug(f"🔧 Sending upstream request to: {upstream_url}")
+            logger.debug(f"🔧 has_function_call: {has_function_call}")
+            logger.debug(f"🔧 Request body contains tools: {bool(body.tools)}")
+            
+            upstream_response = await http_client.post(
+                upstream_url, json=request_body_dict, headers=headers, timeout=app_config.server.timeout
+            )
+            upstream_response.raise_for_status() # If status code is 4xx or 5xx, raise exception
+            
+            response_json = upstream_response.json()
+            logger.debug(f"🔧 Upstream response status code: {upstream_response.status_code}")
+            
+            if has_function_call:
+                content = response_json["choices"][0]["message"]["content"]
+                logger.debug(f"🔧 Complete response content: {repr(content)}")
+                
+                parsed_tools = parse_function_calls_xml(content, GLOBAL_TRIGGER_SIGNAL)
+                logger.debug(f"🔧 XML parsing result: {parsed_tools}")
+                
+                if parsed_tools:
+                    logger.debug(f"🔧 Successfully parsed {len(parsed_tools)} tool calls")
+                    tool_calls = []
+                    for tool in parsed_tools:
+                        tool_call_id = f"call_{uuid.uuid4().hex}"
+                        store_tool_call_mapping(
+                            tool_call_id,
+                            tool["name"],
+                            tool["args"],
+                            f"Calling tool {tool['name']}"
+                        )
+                        tool_calls.append({
+                            "id": tool_call_id,
+                            "type": "function",
+                            "function": {
+                                "name": tool["name"],
+                                "arguments": json.dumps(tool["args"])
+                            }
+                        })
+                    logger.debug(f"🔧 Converted tool_calls: {tool_calls}")
+                    
+                    response_json["choices"][0]["message"] = {
+                        "role": "assistant",
+                        "content": None,
+                        "tool_calls": tool_calls,
+                    }
+                    response_json["choices"][0]["finish_reason"] = "tool_calls"
+                    logger.debug(f"🔧 Function call conversion completed")
+                else:
+                    logger.debug(f"🔧 No tool calls detected, returning original content (including think blocks)")
+            else:
+                logger.debug(f"🔧 No function calls detected or conversion conditions not met")
+            
+            return JSONResponse(content=response_json)
+
+        except httpx.HTTPStatusError as e:
+            logger.error(f"❌ Upstream service response error: status_code={e.response.status_code}")
+            logger.error(f"❌ Upstream error details: {e.response.text}")
+            
+            if e.response.status_code == 400:
+                error_response = {
+                    "error": {
+                        "message": "Invalid request parameters",
+                        "type": "invalid_request_error",
+                        "code": "bad_request"
+                    }
+                }
+            elif e.response.status_code == 401:
+                error_response = {
+                    "error": {
+                        "message": "Authentication failed",
+                        "type": "authentication_error", 
+                        "code": "unauthorized"
+                    }
+                }
+            elif e.response.status_code == 403:
+                error_response = {
+                    "error": {
+                        "message": "Access forbidden",
+                        "type": "permission_error",
+                        "code": "forbidden"
+                    }
+                }
+            elif e.response.status_code == 429:
+                error_response = {
+                    "error": {
+                        "message": "Rate limit exceeded",
+                        "type": "rate_limit_error",
+                        "code": "rate_limit_exceeded"
+                    }
+                }
+            elif e.response.status_code >= 500:
+                error_response = {
+                    "error": {
+                        "message": "Upstream service temporarily unavailable",
+                        "type": "service_error",
+                        "code": "upstream_error"
+                    }
+                }
+            else:
+                error_response = {
+                    "error": {
+                        "message": "Request processing failed",
+                        "type": "api_error",
+                        "code": "unknown_error"
+                    }
+                }
+            
+            return JSONResponse(content=error_response, status_code=e.response.status_code)
+        
+    else:
+        return StreamingResponse(
+            stream_proxy_with_fc_transform(upstream_url, request_body_dict, headers, body.model, has_function_call, GLOBAL_TRIGGER_SIGNAL),
+            media_type="text/event-stream"
+        )
+
+async def stream_proxy_with_fc_transform(url: str, body: dict, headers: dict, model: str, has_fc: bool, trigger_signal: str):
+    """
+    Enhanced streaming proxy, supports dynamic trigger signals, avoids misjudgment within think tags
+    """
+    logger.info(f"📝 Starting streaming response from: {url}")
+    logger.info(f"📝 Function calling enabled: {has_fc}")
+
+    if not has_fc or not trigger_signal:
+        try:
+            async with http_client.stream("POST", url, json=body, headers=headers, timeout=app_config.server.timeout) as response:
+                async for chunk in response.aiter_bytes():
+                    yield chunk
+        except httpx.RemoteProtocolError:
+            logger.debug("🔧 Upstream closed connection prematurely, ending stream response")
+            return
+        return
+# setattr()``
+    detector = StreamingFunctionCallDetector(trigger_signal)
+
+    def _prepare_tool_calls(parsed_tools: List[Dict[str, Any]]):
+        tool_calls = []
+        for i, tool in enumerate(parsed_tools):
+            tool_call_id = f"call_{uuid.uuid4().hex}"
+            store_tool_call_mapping(
+                tool_call_id,
+                tool["name"],
+                tool["args"],
+                f"Calling tool {tool['name']}"
+            )
+            tool_calls.append({
+                "index": i, "id": tool_call_id, "type": "function",
+                "function": { "name": tool["name"], "arguments": json.dumps(tool["args"]) }
+            })
+        return tool_calls
+
+    def _build_tool_call_sse_chunks(parsed_tools: List[Dict[str, Any]], model_id: str) -> List[str]:
+        tool_calls = _prepare_tool_calls(parsed_tools)
+        chunks: List[str] = []
+
+        initial_chunk = {
+            "id": f"chatcmpl-{uuid.uuid4().hex}", "object": "chat.completion.chunk",
+            "created": int(os.path.getmtime(__file__)), "model": model_id,
+            "choices": [{"index": 0, "delta": {"role": "assistant", "content": None, "tool_calls": tool_calls}, "finish_reason": None}],
+        }
+        chunks.append(f"data: {json.dumps(initial_chunk)}\n\n")
+
+
+        final_chunk = {
+             "id": f"chatcmpl-{uuid.uuid4().hex}", "object": "chat.completion.chunk",
+            "created": int(os.path.getmtime(__file__)), "model": model_id,
+            "choices": [{"index": 0, "delta": {}, "finish_reason": "tool_calls"}],
+        }
+        chunks.append(f"data: {json.dumps(final_chunk)}\n\n")
+        chunks.append("data: [DONE]\n\n")
+        return chunks
+
+    try:
+        async with http_client.stream("POST", url, json=body, headers=headers, timeout=app_config.server.timeout) as response:
+            if response.status_code != 200:
+                error_content = await response.aread()
+                logger.error(f"❌ Upstream service stream response error: status_code={response.status_code}")
+                logger.error(f"❌ Upstream error details: {error_content.decode('utf-8', errors='ignore')}")
+                
+                if response.status_code == 401:
+                    error_message = "Authentication failed"
+                elif response.status_code == 403:
+                    error_message = "Access forbidden"
+                elif response.status_code == 429:
+                    error_message = "Rate limit exceeded"
+                elif response.status_code >= 500:
+                    error_message = "Upstream service temporarily unavailable"
+                else:
+                    error_message = "Request processing failed"
+                
+                error_chunk = {"error": {"message": error_message, "type": "upstream_error"}}
+                yield f"data: {json.dumps(error_chunk)}\n\n"
+                yield "data: [DONE]\n\n"
+                return
+
+            async for line in response.aiter_lines():
+                if detector.state == "tool_parsing":
+                    if line.startswith("data:"):
+                        line_data = line[len("data: "):].strip()
+                        if line_data and line_data != "[DONE]":
+                            try:
+                                chunk_json = json.loads(line_data)
+                                delta_content = chunk_json.get("choices", [{}])[0].get("delta", {}).get("content", "") or ""
+                                detector.content_buffer += delta_content
+                                # Early termination: once </function_calls> appears, parse and finish immediately
+                                if "</function_calls>" in detector.content_buffer:
+                                    logger.debug("🔧 Detected </function_calls> in stream, finalizing early...")
+                                    parsed_tools = detector.finalize()
+                                    if parsed_tools:
+                                        logger.debug(f"🔧 Early finalize: parsed {len(parsed_tools)} tool calls")
+                                        for sse in _build_tool_call_sse_chunks(parsed_tools, model):
+                                            yield sse
+                                        return
+                                    else:
+                                        logger.error("❌ Early finalize failed to parse tool calls")
+                                        error_content = "Error: Detected tool use signal but failed to parse function call format"
+                                        error_chunk = { "id": "error-chunk", "choices": [{"delta": {"content": error_content}}]}
+                                        yield f"data: {json.dumps(error_chunk)}\n\n"
+                                        yield "data: [DONE]\n\n"
+                                        return
+                            except (json.JSONDecodeError, IndexError):
+                                pass
+                    continue
+                
+                if line.startswith("data:"):
+                    line_data = line[len("data: "):].strip()
+                    if not line_data or line_data == "[DONE]":
+                        continue
+                    
+                    try:
+                        chunk_json = json.loads(line_data)
+                        delta_content = chunk_json.get("choices", [{}])[0].get("delta", {}).get("content", "") or ""
+                        
+                        if delta_content:
+                            is_detected, content_to_yield = detector.process_chunk(delta_content)
+                            
+                            if content_to_yield:
+                                yield_chunk = {
+                                    "id": f"chatcmpl-passthrough-{uuid.uuid4().hex}",
+                                    "object": "chat.completion.chunk",
+                                    "created": int(os.path.getmtime(__file__)),
+                                    "model": model,
+                                    "choices": [{"index": 0, "delta": {"content": content_to_yield}}]
+                                }
+                                yield f"data: {json.dumps(yield_chunk)}\n\n"
+                            
+                            if is_detected:
+                                # Tool call signal detected, switch to parsing mode
+                                continue
+                    
+                    except (json.JSONDecodeError, IndexError):
+                        yield line + "\n\n"
+
+    except httpx.RequestError as e:
+        logger.error(f"❌ Failed to connect to upstream service: {e}")
+        logger.error(f"❌ Error type: {type(e).__name__}")
+        
+        error_message = "Failed to connect to upstream service"
+        error_chunk = {"error": {"message": error_message, "type": "connection_error"}}
+        yield f"data: {json.dumps(error_chunk)}\n\n"
+        yield "data: [DONE]\n\n"
+        return
+
+    if detector.state == "tool_parsing":
+        logger.debug(f"🔧 Stream ended, starting to parse tool call XML...")
+        parsed_tools = detector.finalize()
+        if parsed_tools:
+            logger.debug(f"🔧 Streaming processing: Successfully parsed {len(parsed_tools)} tool calls")
+            for sse in _build_tool_call_sse_chunks(parsed_tools, model):
+                yield sse
+            return
+        else:
+            logger.error(f"❌ Detected tool call signal but XML parsing failed, buffer content: {detector.content_buffer}")
+            error_content = "Error: Detected tool use signal but failed to parse function call format"
+            error_chunk = { "id": "error-chunk", "choices": [{"delta": {"content": error_content}}]}
+            yield f"data: {json.dumps(error_chunk)}\n\n"
+
+    elif detector.state == "detecting" and detector.content_buffer:
+        # If stream has ended but buffer still has remaining characters insufficient to form signal, output them
+        final_yield_chunk = {
+            "id": f"chatcmpl-finalflush-{uuid.uuid4().hex}", "object": "chat.completion.chunk",
+            "created": int(os.path.getmtime(__file__)), "model": model,
+            "choices": [{"index": 0, "delta": {"content": detector.content_buffer}}]
+        }
+        yield f"data: {json.dumps(final_yield_chunk)}\n\n"
+
+    yield "data: [DONE]\n\n"
+
+
+@app.get("/")
+def read_root():
+    return {
+        "status": "OpenAI Function Call Middleware is running",
+        "config": {
+            "upstream_services_count": len(app_config.upstream_services),
+            "client_keys_count": len(app_config.client_authentication.allowed_keys),
+            "models_count": len(MODEL_TO_SERVICE_MAPPING),
+            "features": {
+                "function_calling": app_config.features.enable_function_calling,
+                "log_level": app_config.features.log_level,
+                "convert_developer_to_system": app_config.features.convert_developer_to_system,
+                "random_trigger": True
+            }
+        }
+    }
+
+@app.get("/v1/models")
+async def list_models(_api_key: str = Depends(verify_api_key)):
+    """List all available models"""
+    visible_models = set()
+    for model_name in MODEL_TO_SERVICE_MAPPING.keys():
+        if ':' in model_name:
+            parts = model_name.split(':', 1)
+            if len(parts) == 2:
+                alias, _ = parts
+                visible_models.add(alias)
+            else:
+                visible_models.add(model_name)
+        else:
+            visible_models.add(model_name)
+
+    models = []
+    for model_id in sorted(visible_models):
+        models.append({
+            "id": model_id,
+            "object": "model",
+            "created": 1677610602,
+            "owned_by": "openai",
+            "permission": [],
+            "root": model_id,
+            "parent": None
+        })
+    
+    return {
+        "object": "list",
+        "data": models
+    }
+
+
+def validate_message_structure(messages: List[Dict[str, Any]]) -> bool:
+    """Validate if message structure meets requirements"""
+    try:
+        valid_roles = ["system", "user", "assistant", "tool"]
+        if not app_config.features.convert_developer_to_system:
+            valid_roles.append("developer")
+        
+        for i, msg in enumerate(messages):
+            if "role" not in msg:
+                logger.error(f"❌ Message {i} missing role field")
+                return False
+            
+            if msg["role"] not in valid_roles:
+                logger.error(f"❌ Invalid role value for message {i}: {msg['role']}")
+                return False
+            
+            if msg["role"] == "tool":
+                if "tool_call_id" not in msg:
+                    logger.error(f"❌ Tool message {i} missing tool_call_id field")
+                    return False
+            
+            content = msg.get("content")
+            content_info = ""
+            if content:
+                if isinstance(content, str):
+                    content_info = f", content=text({len(content)} chars)"
+                elif isinstance(content, list):
+                    text_parts = [item for item in content if isinstance(item, dict) and item.get('type') == 'text']
+                    image_parts = [item for item in content if isinstance(item, dict) and item.get('type') == 'image_url']
+                    content_info = f", content=multimodal(text={len(text_parts)}, images={len(image_parts)})"
+                else:
+                    content_info = f", content={type(content).__name__}"
+            else:
+                content_info = ", content=empty"
+            
+            logger.debug(f"✅ Message {i} validation passed: role={msg['role']}{content_info}")
+        
+        logger.debug(f"✅ All messages validated successfully, total {len(messages)} messages")
+        return True
+    except Exception as e:
+        logger.error(f"❌ Message validation exception: {e}")
+        return False
+
+def safe_process_tool_choice(tool_choice) -> str:
+    """Safely process tool_choice field to avoid type errors"""
+    try:
+        if tool_choice is None:
+            return ""
+        
+        if isinstance(tool_choice, str):
+            if tool_choice == "none":
+                return "\n\n**IMPORTANT:** You are prohibited from using any tools in this round. Please respond like a normal chat assistant and answer the user's question directly."
+            else:
+                logger.debug(f"🔧 Unknown tool_choice string value: {tool_choice}")
+                return ""
+        
+        elif hasattr(tool_choice, 'function') and hasattr(tool_choice.function, 'name'):
+            required_tool_name = tool_choice.function.name
+            return f"\n\n**IMPORTANT:** In this round, you must use ONLY the tool named `{required_tool_name}`. Generate the necessary parameters and output in the specified XML format."
+        
+        else:
+            logger.debug(f"🔧 Unsupported tool_choice type: {type(tool_choice)}")
+            return ""
+    
+    except Exception as e:
+        logger.error(f"❌ Error processing tool_choice: {e}")
+        return ""
+
+if __name__ == "__main__":
+    import uvicorn
+    logger.info(f"🚀 Starting server on {app_config.server.host}:{app_config.server.port}")
+    logger.info(f"⏱️  Request timeout: {app_config.server.timeout} seconds")
+    
+    uvicorn.run(
+        app,
+        host=app_config.server.host,
+        port=app_config.server.port,
+        log_level=app_config.features.log_level.lower() if app_config.features.log_level != "DISABLED" else "critical"
+    )

requirements.txt

@@ -0,0 +1,5 @@
+fastapi
+uvicorn
+httpx
+pydantic
+pyyaml

scripts/__init__.py

@@ -0,0 +1 @@
+

scripts/generate_config.py

@@ -0,0 +1,126 @@
+#!/usr/bin/env python3
+"""基于环境变量生成 Toolify 配置文件。"""
+
+import os
+import sys
+from pathlib import Path
+from typing import Any, List
+
+import yaml
+
+
+def _str_to_bool(value: str | None, default: bool) -> bool:
+    """将字符串环境变量解析为布尔值。"""
+    if value is None:
+        return default
+    normalized = value.strip().lower()
+    if normalized in {"1", "true", "yes", "y", "on"}:
+        return True
+    if normalized in {"0", "false", "no", "n", "off"}:
+        return False
+    return default
+
+
+def _parse_list(env_value: str | None) -> List[str]:
+    """将逗号分隔的环境变量解析为字符串列表。"""
+    if not env_value:
+        return []
+    return [item.strip() for item in env_value.split(",") if item.strip()]
+
+
+def _load_config_from_env() -> Any:
+    """尝试从直接提供的 YAML 或 JSON 环境变量加载配置。"""
+    direct_yaml = os.getenv("TOOLIFY_CONFIG_YAML")
+    if direct_yaml:
+        return yaml.safe_load(direct_yaml)
+
+    direct_json = os.getenv("TOOLIFY_CONFIG_JSON")
+    if direct_json:
+        return yaml.safe_load(direct_json)
+
+    return None
+
+
+def _build_default_config() -> dict:
+    """根据基础环境变量构造最小可用配置。"""
+    openai_api_key = os.getenv("OPENAI_API_KEY")
+    allowed_keys = _parse_list(os.getenv("CLIENT_ALLOWED_KEYS"))
+
+    if not openai_api_key:
+        raise RuntimeError("缺少 OPENAI_API_KEY 环境变量，无法生成配置。")
+
+    if not allowed_keys:
+        raise RuntimeError("缺少 CLIENT_ALLOWED_KEYS 环境变量，无法生成配置。")
+
+    base_url = os.getenv("OPENAI_BASE_URL", "https://api.openai.com/v1").rstrip("/")
+    models = _parse_list(os.getenv("OPENAI_MODELS", "gpt-4o-mini,gpt-4o"))
+    if not models:
+        raise RuntimeError("OPENAI_MODELS 环境变量解析为空，请至少指定一个模型。")
+
+    server_port = int(os.getenv("SERVER_PORT", os.getenv("PORT", "8000")))
+    server_host = os.getenv("SERVER_HOST", "0.0.0.0")
+    server_timeout = int(os.getenv("SERVER_TIMEOUT", "180"))
+
+    features_log_level = os.getenv("FEATURE_LOG_LEVEL", "INFO").upper()
+
+    prompt_template = os.getenv("FEATURE_PROMPT_TEMPLATE")
+    if prompt_template:
+        prompt_template = prompt_template.replace("\\n", "\n")
+
+    return {
+        "server": {
+            "port": server_port,
+            "host": server_host,
+            "timeout": server_timeout,
+        },
+        "upstream_services": [
+            {
+                "name": os.getenv("OPENAI_NAME", "openai"),
+                "base_url": base_url,
+                "api_key": openai_api_key,
+                "description": os.getenv("OPENAI_DESCRIPTION", "OpenAI Service"),
+                "is_default": True,
+                "models": models,
+            }
+        ],
+        "client_authentication": {
+            "allowed_keys": allowed_keys,
+        },
+        "features": {
+            "enable_function_calling": _str_to_bool(os.getenv("FEATURE_ENABLE_FUNCTION_CALLING"), True),
+            "log_level": features_log_level if features_log_level else "INFO",
+            "convert_developer_to_system": _str_to_bool(os.getenv("FEATURE_CONVERT_DEVELOPER_TO_SYSTEM"), True),
+            "prompt_template": prompt_template,
+            "key_passthrough": _str_to_bool(os.getenv("FEATURE_KEY_PASSTHROUGH"), False),
+            "model_passthrough": _str_to_bool(os.getenv("FEATURE_MODEL_PASSTHROUGH"), False),
+        },
+    }
+
+
+def main() -> None:
+    """入口函数：生成配置文件。"""
+    config_path = Path(os.getenv("CONFIG_PATH", "/app/config.yaml"))
+
+    if config_path.exists():
+        print(f"配置文件已存在，跳过生成: {config_path}")
+        return
+
+    config_obj = _load_config_from_env()
+    if config_obj is None:
+        config_obj = _build_default_config()
+
+    if not isinstance(config_obj, dict):
+        raise ValueError("环境变量提供的配置必须是字典结构。")
+
+    config_path.parent.mkdir(parents=True, exist_ok=True)
+    config_yaml = yaml.safe_dump(config_obj, sort_keys=False, allow_unicode=True)
+    config_path.write_text(config_yaml, encoding="utf-8")
+    print(f"已生成配置文件: {config_path}")
+
+
+if __name__ == "__main__":
+    try:
+        main()
+    except Exception as exc:  # noqa: BLE001
+        print(f"生成配置文件时出现错误: {exc}", file=sys.stderr)
+        sys.exit(1)