The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from external contributors which could contain malicious instructions designed to manipulate the AI during the release process.
Ingestion points: The skill fetches commit messages and pull request metadata via gh api and gh pr list commands as shown in the SKILL.md workflows.
Boundary markers: There are no explicit delimiters or instructions to the AI agent to ignore potentially malicious embedded directives in the fetched text.
Capability inventory: The skill leverages high-privilege capabilities including shell command execution (Bash) and file writing (Write) which could be misused if an injection is successful.
Sanitization: No explicit sanitization or filtering of the external commit and PR data is performed before it is used to generate release artifacts.
[COMMAND_EXECUTION]: The skill makes extensive use of the Bash tool to execute gh CLI commands and Node.js scripts for repository management, building artifacts, and coordinating deployments.
[EXTERNAL_DOWNLOADS]: The skill utilizes npx to dynamically retrieve and execute the claude-flow utility and its associated swarm modules from the npm registry during the release pipeline.

github-release-management