The Agent Skills Directory

[SAFE]: The skill functions as a procedural guide for development workflows. It instructs the agent to execute standard linting and formatting tools (e.g., eslint, prettier, ruff, black) and run test suites expected to be found within a project's existing environment.
[PROMPT_INJECTION]: The skill relies on information retrieved from project-specific files like CLAUDE.md, package.json, and Gemfile to determine which commands to execute. This constitutes an indirect prompt injection surface where a maliciously modified project file could attempt to influence the agent's actions during the finalization process.
Ingestion points: CLAUDE.md, package.json, Gemfile (referenced in SKILL.md)
Boundary markers: Absent
Capability inventory: Subprocess execution (linting/testing) and file system writes (fixing code issues)
Sanitization: Absent

finalize