The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified. The skill performs legitimate repository maintenance tasks using standard tools.
[COMMAND_EXECUTION]: The skill executes local shell commands such as find and grep through scripts/main.cjs and documented workflows in SKILL.md. These commands are used for file discovery and text searching within the project directory. The implementation in main.cjs uses execSync with shell: false, which is a secure method for executing system commands.
[DATA_EXFILTRATION]: Execution telemetry is recorded locally to context/runtime/tool-events.jsonl via hooks/post-execute.cjs. No network operations or external data transmissions were found.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it incorporates content from repository files into its reports.
Ingestion points: Output from find and grep commands on local repository files.
Boundary markers: None; findings are placed directly into markdown reports.
Capability inventory: Uses Bash and execSync for local file system auditing.
Sanitization: None, though output is filtered for specific patterns and truncated to 20 results per category.

gap-detection