The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the langwatch CLI to perform diagnostic tasks. Commands such as langwatch trace search, langwatch trace get, and langwatch monitor list are used to retrieve technical data from the platform for analysis.
[EXTERNAL_DOWNLOADS]: The skill references documentation hosted at langwatch.ai. It retrieves guides and reference files (e.g., llms.txt) via HTTP to provide the agent with current context on SDK integrations and CLI usage.
[REMOTE_CODE_EXECUTION]: The skill includes a procedure to run npx langwatch report to submit bug reports or feedback. This command executes the vendor's official reporting tool and requires the --user-approved flag, ensuring user oversight before any data is sent.
[DATA_EXFILTRATION]: The skill is designed to fetch and display trace data, which may contain sensitive LLM inputs and outputs, from the LangWatch production environment. This access is local to the agent's session and serves the primary purpose of the skill: identifying the cause of application errors.

debug-with-langwatch