Spaces:
Running
Running
| <html lang="en"> | |
| <head> | |
| <meta charset="utf-8"> | |
| <meta name="viewport" content="width=device-width, initial-scale=1"> | |
| <title>The Agent Reliability Stack — Tour</title> | |
| <style> | |
| :root { | |
| --bg: #1a1a1a; | |
| --fg: #e8e6e1; | |
| --muted: #9a9690; | |
| --accent: #d4a853; | |
| --card: #232323; | |
| --border: #353535; | |
| } | |
| * { box-sizing: border-box; } | |
| body { | |
| font: 16px/1.55 -apple-system, BlinkMacSystemFont, 'SF Pro Text', sans-serif; | |
| background: var(--bg); | |
| color: var(--fg); | |
| margin: 0; | |
| padding: 2rem 1.25rem; | |
| } | |
| main { max-width: 880px; margin: 0 auto; } | |
| h1 { color: var(--accent); font-size: 2.1rem; margin: 0 0 0.5rem; } | |
| h2 { color: var(--accent); font-size: 1.4rem; margin: 2.2rem 0 0.6rem; } | |
| .lede { color: var(--muted); font-size: 1.05rem; max-width: 680px; } | |
| .pipeline { | |
| display: flex; gap: 0.4rem; flex-wrap: wrap; align-items: center; | |
| font-family: ui-monospace, SF Mono, Menlo, monospace; | |
| background: var(--card); border: 1px solid var(--border); | |
| padding: 0.9rem 1.1rem; border-radius: 8px; margin: 1.2rem 0; | |
| } | |
| .pipeline span { color: var(--accent); font-weight: 600; } | |
| .pipeline span.arrow { color: var(--muted); font-weight: 400; } | |
| .grid { | |
| display: grid; gap: 1rem; | |
| grid-template-columns: repeat(auto-fit, minmax(260px, 1fr)); | |
| margin: 1.2rem 0; | |
| } | |
| .card { | |
| background: var(--card); border: 1px solid var(--border); | |
| padding: 1.1rem 1.2rem; border-radius: 10px; | |
| } | |
| .card h3 { margin: 0 0 0.4rem; color: var(--accent); font-size: 1.1rem; } | |
| .card .one-liner { color: var(--muted); font-size: 0.95rem; margin-bottom: 0.5rem; } | |
| .card a { color: var(--fg); text-decoration: none; border-bottom: 1px solid var(--border); } | |
| .card a:hover { border-color: var(--accent); } | |
| code { background: var(--card); padding: 0.1rem 0.4rem; border-radius: 4px; font-size: 0.9em; } | |
| pre { background: var(--card); border: 1px solid var(--border); padding: 1rem; border-radius: 8px; overflow-x: auto; } | |
| a { color: var(--accent); } | |
| footer { color: var(--muted); margin-top: 3rem; padding-top: 1.5rem; border-top: 1px solid var(--border); font-size: 0.9rem; } | |
| </style> | |
| </head> | |
| <body> | |
| <main> | |
| <h1>The Agent Reliability Stack</h1> | |
| <p class="lede">Five small, focused libraries that fix the boring problems every long-running AI agent eventually hits. Pure Python, zero runtime deps. BYO LLM.</p> | |
| <div class="pipeline"> | |
| <span>fit</span><span class="arrow">→</span> | |
| <span>guard</span><span class="arrow">→</span> | |
| <span>snap</span><span class="arrow">→</span> | |
| <span>vet</span><span class="arrow">→</span> | |
| <span>cast</span> | |
| </div> | |
| <h2>The five libraries</h2> | |
| <div class="grid"> | |
| <div class="card"> | |
| <h3>🪟 fit</h3> | |
| <div class="one-liner">Fit a chat history into a token budget.</div> | |
| <p>Drop-oldest, drop-middle, or priority-based truncation. Pluggable tokenizers. Preserves system + last-N messages.</p> | |
| <a href="https://pypi.org/project/agentfit-py/">PyPI</a> · <a href="https://huggingface.co/spaces/mukunda1729/token-counter">Token counter</a> | |
| </div> | |
| <div class="card"> | |
| <h3>🛡️ guard</h3> | |
| <div class="one-liner">Network egress firewall for agent tools.</div> | |
| <p>Declarative allow/deny lists for any URL the agent tries to fetch. Throws on violation, before the request leaves your process.</p> | |
| <a href="https://pypi.org/project/agentguard-firewall/">PyPI</a> | |
| </div> | |
| <div class="card"> | |
| <h3>📸 snap</h3> | |
| <div class="one-liner">Diff two tool-call traces.</div> | |
| <p>Catch silent regressions in agent pipelines: same input, different tools, different output. Snapshot tests for agents.</p> | |
| <a href="https://pypi.org/project/agentsnap-py/">PyPI</a> · <a href="https://huggingface.co/datasets/mukunda1729/agent-trace-samples">Sample traces</a> | |
| </div> | |
| <div class="card"> | |
| <h3>✅ vet</h3> | |
| <div class="one-liner">Validate tool-call args before execution.</div> | |
| <p>Wraps any tool. Returns LLM-friendly retry hints when args are wrong, so the model can self-correct.</p> | |
| <a href="https://pypi.org/project/agentvet-py/">PyPI</a> | |
| </div> | |
| <div class="card"> | |
| <h3>🎯 cast</h3> | |
| <div class="one-liner">Extract JSON from messy LLM output.</div> | |
| <p>Tolerant extractor that handles fenced blocks, prose-wrapped JSON, refusals. Validates against a shape.</p> | |
| <a href="https://pypi.org/project/agentcast-py/">PyPI</a> · <a href="https://huggingface.co/spaces/mukunda1729/json-extractor">JSON extractor</a> | |
| </div> | |
| </div> | |
| <h2>Install</h2> | |
| <pre>pip install mk-agentkit # all five | |
| pip install agentfit-py # individual | |
| npm install @mukundakatta/agentkit # all five | |
| npm install @mukundakatta/agentfit # individual</pre> | |
| <h2>Try the live demos</h2> | |
| <ul> | |
| <li><a href="https://huggingface.co/spaces/mukunda1729/agent-stack-demo">agent-stack-demo</a> — all 5 libs in one Space</li> | |
| <li><a href="https://huggingface.co/spaces/mukunda1729/token-counter">token-counter</a> — across 5 model families</li> | |
| <li><a href="https://huggingface.co/spaces/mukunda1729/json-extractor">json-extractor</a> — for messy LLM text</li> | |
| <li><a href="https://huggingface.co/spaces/mukunda1729/pii-redactor">pii-redactor</a> — find emails / secrets / IDs</li> | |
| <li><a href="https://huggingface.co/spaces/mukunda1729/prompt-injection-detector">prompt-injection-detector</a> — heuristic scanner</li> | |
| <li><a href="https://huggingface.co/spaces/mukunda1729/mcp-config-validator">mcp-config-validator</a> — sanity-check MCP configs</li> | |
| </ul> | |
| <h2>Datasets</h2> | |
| <p>13 public datasets covering agent traces, prompt injections, MCP configs, hallucination cases, and more. <a href="https://huggingface.co/mukunda1729">Browse all on HuggingFace →</a></p> | |
| <footer> | |
| Built by <a href="https://github.com/MukundaKatta">Mukunda Katta</a> · MIT licensed across the board · <a href="https://mukundakatta.github.io/agent-stack/">Landing page</a> | |
| </footer> | |
| </main> | |
| </body> | |
| </html> | |