The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes various local development commands including uv, npm, npx, pytest, and git. This is the intended primary purpose of the skill to facilitate code verification within a development environment.
[PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill ingests and processes output from external tools which could potentially contain adversarial instructions embedded in logs or test failure messages.
Ingestion points: The skill captures and processes output from uv build, npm run build, npx tsc, pyright, ruff, pytest, grep, and git across all verification phases.
Boundary markers: Absent. There are no instructions or delimiters provided to the agent to distinguish between tool output and instructions, nor warnings to ignore embedded content.
Capability inventory: The skill leverages subprocess execution to run project-specific build systems, test runners, and file system searches.
Sanitization: Absent. Data from external tool outputs is interpolated directly into the agent's context without sanitization or validation.

verification-loop