The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses uv run to execute a local utility script, scripts/parse_errors.py, for analyzing compiler output. This is a standard and safe use of local tooling.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data in the form of error logs. While this creates a theoretical attack surface, the use of a script with specific regex filters to process this data before it reaches the agent's main context significantly reduces the risk of the agent misinterpreting malicious instructions embedded in the logs.
Ingestion points: error_log.txt (via --file), --message arguments, and standard input (stdin) in scripts/parse_errors.py.
Boundary markers: No explicit markers or delimiters are defined in the instructions for the log content.
Capability inventory: The skill can read local files and execute the provided Python scripts.
Sanitization: The Python script uses regular expressions to extract specific error patterns, providing a level of structural validation.
[SAFE]: References to external documentation files using relative paths (e.g., ../../docs/board-support/uno-r4-family.md) are consistent with a structured monorepo and do not represent malicious path traversal.

error-message-explainer