The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches documentation and scenario indices from official vendor URLs (e.g., https://langwatch.ai/docs/llms.txt) to provide instructions and context for debugging tasks.\n- [COMMAND_EXECUTION]: Utilizes the langwatch CLI for experiment management operations, including listing runs, filtering results, and inspecting evaluators.\n- [REMOTE_CODE_EXECUTION]: Instructs the agent to run npx langwatch report, which downloads and executes the vendor-supplied langwatch package from the registry.\n- [DATA_EXFILTRATION]: Provides a mechanism to send diagnostic data and transcripts to the vendor's reporting endpoint via the report command. The instructions state that the tool redacts secrets and personal data locally before transmission.\n- [PROMPT_INJECTION]: The skill ingests external content from documentation URLs which is then processed by the agent. Although no explicit boundary markers or sanitization steps are defined for this input, the data originates from the official vendor site and is used to drive interactions within the scope of the langwatch CLI tool.

eval-triage