The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon user-supplied error logs and environment data to trigger diagnostic actions.
Ingestion points: The skill instructions prompt the agent to process the "exact error, SDK version, and Langfuse base URL" from the user's failing environment.
Boundary markers: There are no explicit delimiters for the input data, though the instructions provide clear guidance on output sanitization.
Capability inventory: The skill utilizes Bash (curl) for network testing, npm for dependency management, and Read/Grep for local file analysis.
Sanitization: The instructions explicitly warn the agent: "Do not include API keys, authorization headers, or raw customer prompts in the report," and use shell variable expansion to mask secrets in diagnostic output.
[COMMAND_EXECUTION]: The skill uses shell commands to diagnose connectivity and verify local environment state.
Network Diagnostics: Executes curl health checks against the Langfuse API and performs nslookup for DNS verification.
Environment Inspection: Uses npm list and pip show to verify installed SDK versions and docker ps to check local container status.
[EXTERNAL_DOWNLOADS]: The troubleshooting steps recommend updating or installing Langfuse SDKs from the official npm registry.
Resource Fetching: Suggested fixes involve running npm install @langfuse/client@latest @langfuse/tracing@latest @langfuse/otel@latest to resolve version mismatches.

langfuse-common-errors