Metadata-Version: 2.4
Name: hlido-langchain
Version: 0.1.0
Summary: LangChain tools for Hlido — vet an AI agent against its independent, evidence-backed trust score before delegating to it.
Author-email: Hlido <hello@hlido.eu>
License: MIT
Project-URL: Homepage, https://hlido.eu
Project-URL: Documentation, https://hlido.eu/api/
Project-URL: Source, https://github.com/ankitkapur1992-hlido/hlido-public/tree/main/integrations/hlido-langchain
Project-URL: Issues, https://github.com/ankitkapur1992-hlido/hlido-public/issues
Keywords: ai,agents,trust,langchain,evaluation,reliability,hlido
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Requires-Python: >=3.9
Description-Content-Type: text/markdown
Requires-Dist: hlido-trust[langchain]>=0.1.0

# hlido-langchain

LangChain tools for [Hlido](https://hlido.eu) — independent, evidence-backed trust
scores for AI agents. Vet an agent or tool against its independent review **before
delegating to it**.

```bash
pip install hlido-langchain
```

```python
from hlido_langchain import get_tools

tools = get_tools()          # [hlido_trust_check, hlido_recommend]
llm_with_tools = llm.bind_tools(tools)
```

No API key is required for trust checks.

- `hlido_trust_check(slug, min_score=70)` — PASS/FAIL gate + score, tier, red flags, evidence URL.
- `hlido_recommend(need, k=1)` — Hlido-vetted agents for a free-text need, ranked.

Thin wrapper over [`hlido-trust`](https://pypi.org/project/hlido-trust/); the tools live
in `hlido_trust.langchain_tool`. MIT.
