Add Litmus to AI & LLM Testing

[rylinjames]

Adding Litmus — an open-source Python CLI that records and replays AI agent LLM calls deterministically.

What makes it unique: Most LLM testing tools focus on prompt evaluation or observability. Litmus captures the full execution trace (every API call) and replays it deterministically — like VCR/cassette testing but for multi-step AI agents. It also supports fault injection (inject timeouts, errors, refusals) and reliability scoring for CI gating.

• MIT licensed, 47 tests, supports 14+ LLM providers
• Zero code changes needed — wraps any Python agent
• pip install litmus-trace