The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of the Bash tool and python3 to execute diagnostic scripts. It uses utilities such as find, grep, wc, sed, and awk to perform file system analysis, line counting, and pattern matching across the project directory.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted project files which serves as a vulnerability surface. Malicious content within these files could potentially influence the agent's behavior or exploit the command execution environment.
Ingestion points: The skill reads CLAUDE.md, .claude/rules/*.md, CATALOG.md, session_*.md, and .claude/settings.json (via the Read, Grep, and Bash tools).
Boundary markers: No specific delimiters or "ignore instructions" warnings are used when reading these files into the agent's context.
Capability inventory: The skill possesses significant capabilities, including arbitrary Bash command execution and python3 script execution across all scripts in SKILL_detail.md.
Sanitization: Content from the ingested files is used in script logic without explicit sanitization or validation, potentially allowing for command injection if filenames or file contents are crafted maliciously.

harness-doctor