The Agent Skills Directory

[COMMAND_EXECUTION]: The find-polluter.sh script executes npm test on files discovered through a user-specified search pattern. This is a standard utility function but involves executing shell commands based on file system state.
[DATA_EXPOSURE]: Examples in SKILL.md (Phase 1, Step 4) demonstrate the use of sensitive macOS diagnostic commands (security list-keychains, security find-identity) to troubleshoot code signing issues.
[PROMPT_INJECTION]: The skill's core function involves processing and acting upon untrusted data such as error messages, stack traces, and build logs. This presents an indirect prompt injection surface.
Ingestion points: Systematic processing of error logs and terminal outputs in Phase 1 (Root Cause Investigation).
Boundary markers: The instructions emphasize reading stack traces "completely" but do not define specific delimiters to isolate log data from instructions.
Capability inventory: The skill directs the agent to execute shell commands, read files, and perform git operations during investigation.
Sanitization: No mechanisms are described for sanitizing or escaping malicious instructions that might be embedded in the processed logs or data flows.

systematic-debugging