The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill performs an audit of untrusted repository data, including README files, git commit history, and issue tickets, to generate its output. This creates a vulnerability surface where malicious content embedded in these files could potentially manipulate the agent's behavior during the generation process.
Ingestion points: README.md, git log, docs/specs/, issues/, tickets/, and various project configuration files.
Boundary markers: Not present.
Capability inventory: Writing to files (scripts/dev.sh, docs/verification/manual-test-guide.md) and identifying/executing shell commands.
Sanitization: No explicit filtering or sanitization of ingested content is defined.
[COMMAND_EXECUTION]: The skill is designed to detect and recommend the execution of common development commands such as npm run dev, docker compose up, and pytest to verify implementation.
[DYNAMIC_EXECUTION]: The skill programmatically creates a shell script (scripts/dev.sh) based on the detected project environment (e.g., Vite, Next.js, FastAPI) to facilitate service startup.

verify-manual-after-implementation