The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands using gh, node, and dotnet to perform repository inventory, run internal triage tests, and validate the project source code. These commands are necessary for the skill's stated purpose of repository triage.
[EXTERNAL_DOWNLOADS]: The skill fetches triage artifacts and patches from GitHub's infrastructure using gh run download and gh pr diff. As these are sourced from a well-known and trusted service, the downloads are categorized as safe.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it reads and processes user-generated content from GitHub issues and pull requests.
Ingestion points: Content is retrieved from the openclaw/openclaw-windows-node repository using GitHub CLI commands and artifact downloads as described in SKILL.md.
Boundary markers: There are no explicit instructions for using delimiters or boundary markers when the agent processes issue and pull request text.
Capability inventory: The skill can execute local commands (node, dotnet, gh), write various files to the session directory, and interface with an interactive dashboard tool.
Sanitization: The skill instructions do not specify any sanitization or validation logic for the incoming data from GitHub.

global-repo-triage