The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes output from external tools including build logs, test results, and git diffs, which serves as a potential ingestion point for indirect prompt injections if the underlying codebase contains malicious content intended to influence the agent.
Ingestion points: The agent reads and reports output from npm, pnpm, npx, pyright, ruff, and git commands executed in the local environment (SKILL.md).
Boundary markers: The instructions do not implement specific delimiters or 'ignore' directives to isolate external tool output from the agent's core instruction context.
Capability inventory: The skill utilizes shell execution capabilities for build systems, linters, and version control systems across various files.
Sanitization: The skill does not perform sanitization or validation of the tool output before it is summarized into the final verification report.

verification-loop