# Judge whether the agent took the right steps, not only gave the right answer (slim)

> Judge whether the agent took the right steps, not only gave the right answer: 2 companies fully do it, 10 with a limit, 20 carry the tag. Every listing in Agent observability & evals with the reason and a source.

- Full: https://www.anchorterminal.com/niches/agent-observability.agent-trajectory-eval/index.md (~1,800 tokens) · this version ~380 tokens · JSON https://www.anchorterminal.com/niches/agent-observability.agent-trajectory-eval/index.json · canonical https://www.anchorterminal.com/niches/agent-observability.agent-trajectory-eval/
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · MCP: https://www.anchorterminal.com/mcp · Updated: 2026-11-15
- Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data)

Two or three companies' docs show a full fit. Full fit: 2 companies, with a limit: 10, tagged: 20. Weight 5, 6 accounts.

Done when: A score per trajectory that names the wrong or missing tool call or the turn where the run went off, reproducible on the same trace, with a clear result when the agent answered without calling any tool.
yes: langwatch (Reasoning Engine B.V. (doing business as LangWatch))
yes: opik (Comet ML, Inc.)
caveat: arize-phoenix (Arize AI, Inc.)
caveat: braintrust (Braintrust Data, Inc.)
caveat: deepeval (Confident AI, Inc.)
caveat: galileo (Galileo Technologies, Inc.)
caveat: gauge-cli (Gauge)
caveat: honeyhive (HoneyHive Inc.)
caveat: laminar (LMNR AI, Inc.)
caveat: langsmith (LangChain, Inc.)
caveat: mlflow-tracing (MLflow Project, a Series of LF Projects, LLC)
caveat: overmind (Overmind Technology Ltd (England and Wales, company 16585996))
unchecked: pydantic-logfire (Pydantic Services Inc.)

Data: https://www.anchorterminal.com/api/v1/use-cases/agent-observability.agent-trajectory-eval.json
