Hugging Face's logo

Spaces:
mukunda1729
/
agent-stack-tour
Running

App Files Community
agent-stack-tour
File size: 6,096 Bytes
86f42a9
cbaa5f0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
86f42a9
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
<!doctype html>
<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>