Legal Consulting Bundle

What to consider before installing: - The code will run a local FastAPI web app and (optionally) call an external LLM service (DeepSeek) if you provide DEEPSEEK_API_KEY. The SKILL.md/manifest do not list this env var even though main.py reads it — do not set an API key unless you trust DeepSeek and understand that user queries (and included knowledge/context) will be transmitted to that external service. - .env.example referenced by the README/SKILL.md is missing from the package. Inspect any .env you create carefully — do not paste secrets or sensitive personal data into the service while testing. - The knowledge mapping in code references some law files that are not included; expect some features to use fallback templates rather than a full LLM analysis. Test in 'template mode' (no DEEPSEEK_API_KEY) to avoid external calls; template mode returns canned responses. - Run the app in an isolated environment (container or VM) and bind the server to localhost only if you do not want it exposed on the network. Review the code paths that call httpx.post and any other network calls before providing credentials. - If you need this functionality but are not comfortable with the unknown external LLM, consider removing/modifying the external call (call_llm) or replacing it with a vetted LLM provider and explicitly document required env vars. - Summary action items: inspect main.py, do not set DEEPSEEK_API_KEY unless you accept data sharing, create a local .env for safe testing, and run in an isolated environment.