The Agent Skills Directory

[COMMAND_EXECUTION]: Executes a comprehensive suite of local development tools across multiple ecosystems (Node.js, Python, Go, Rust, Ruby, Java, Scala, Flutter) to verify build integrity and test success.
[COMMAND_EXECUTION]: Records execution metadata (telemetry) to a file named skill-telemetry.md within the ~/.claude/projects/ directory to support skill evolution pipelines.
[DATA_EXFILTRATION]: Performs a local security scan on tracked files and environment configurations to identify potentially exposed API keys, private keys, or tokens. The results are reported to the user as part of the readiness summary and are not transmitted externally.
[PROMPT_INJECTION]: Processes external project data which creates a surface for indirect prompt injection.
Ingestion points: Reads configuration from CLAUDE.md, package.json, and pyproject.toml; parses git commit logs and pull request descriptions via gh pr view.
Boundary markers: None; the skill does not use specific delimiters or instructions to ignore embedded commands within the processed data.
Capability inventory: The skill has broad shell execution permissions for project tools and local file write access to the agent's project memory directory.
Sanitization: No sanitization or strict validation is applied to the ingested text before it is evaluated against project conventions.

preflight