The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill is configured to use npx ruv-swarm within several lifecycle hooks (pre_task, post_edit, post_task, notification) to download and execute tools from the npm registry. It also performs npm install operations within the build and validation steps.
[REMOTE_CODE_EXECUTION]: The skill downloads and executes external code from the npm registry via npx and npm install during the release coordination process.
[COMMAND_EXECUTION]: The skill makes extensive use of the Bash tool to perform file system operations, clone repositories, and interact with the GitHub API using the gh command-line interface.
[DYNAMIC_EXECUTION]: The release validation workflow involves the dynamic installation of dependencies followed by the execution of tests and build scripts (npm test, npm run build), which runs code generated or resolved at runtime.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: The agent reads and interprets content from external files such as package.json, CHANGELOG.md, and the output of various test suites and build logs.
Boundary markers: The instructions do not include explicit delimiters or guardrails to prevent the agent from following malicious instructions that might be embedded in the processed repository files or tool outputs.
Capability inventory: The skill has broad capabilities, including writing to the local file system, executing shell commands, and modifying GitHub repositories via push and pull request tools.
Sanitization: There is no evidence of content sanitization or validation for the data ingested from the repository before it is processed by the agent.

agent-release-manager