The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes several shell commands to audit the project and generate release notes.
Runs project lifecycle scripts: npm test, npm run lint, and npm run build as part of the readiness checklist.
Executes security audit tools: gitleaks detect --source . and trufflehog git file://. to check for sensitive data leakage.
Uses various git commands (git grep, git status, git log, git diff, git add, git commit) to inspect repository state, extract changes for release notes, and commit generated changesets.
[EXTERNAL_DOWNLOADS]: The skill instructs the user to install external packages from the npm registry.
Recommends installation of @changesets/cli and @changesets/changelog-github as prerequisites.
Uses npx to execute tools like changeset and skills during the auditing and initialization phases.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which could potentially contain malicious instructions designed to influence the AI's behavior.
Ingestion points: Reads content from README.md, Git commit messages (git log), and repository configuration files during the audit and summary generation steps.
Boundary markers: No explicit delimiter-based isolation or "ignore embedded instructions" directives are used when passing project content to sub-agents for review (Step 3) or when summarizing commits.
Capability inventory: The skill has the ability to execute shell commands, read/write files, and spawn sub-agents using the run-agents skill.
Sanitization: There is no evidence of sanitization or filtering of external content (like commit messages or README text) before it is processed by the AI logic.

pre-release