The Agent Skills Directory

[PERSISTENCE_MECHANISMS]: The skill provides commands to modify shell profile files (~/.bashrc, ~/.zshrc) to permanently store the LANGSMITH_API_KEY environment variable, creating a persistent mechanism for sensitive data.
[CREDENTIALS_UNSAFE]: The skill includes instructions to execute 'echo $LANGSMITH_API_KEY', which exposes sensitive API credentials in plain text within the agent's output and logs.
[REMOTE_CODE_EXECUTION]: The skill requires the installation of the 'langsmith-fetch' package via pip, introducing external code into the execution environment that runs with user-level privileges.
[PROMPT_INJECTION]: The skill ingests and analyzes external execution traces and tool outputs, creating a vulnerability to indirect prompt injection attacks where malicious data in the traces could influence the agent's behavior.
Ingestion points: Data fetched via 'langsmith-fetch traces' and 'langsmith-fetch trace' commands.
Boundary markers: Absent; no instructions are provided to isolate or disregard potentially malicious instructions within the fetched logs.
Capability inventory: Execution of shell commands, directory creation, and file system writing.
Sanitization: No sanitization or validation of the fetched trace data is performed before analysis.
[COMMAND_EXECUTION]: The skill relies on shell commands for its core functionality, including the interpolation of user-provided trace IDs, which could lead to command injection if the input is malicious.
[METADATA_POISONING]: There is a discrepancy between the stated author in the vendor context and the metadata within the skill files, which may obscure the provenance and safety of the associated code.

langsmith-fetch