The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes multiple gh CLI commands (gh label create, gh issue create, gh issue edit, gh issue comment, gh issue close, etc.) to manage project state on GitHub. It also runs environment checks such as gh auth status and gh repo view to ensure the tool is correctly configured.
[EXTERNAL_DOWNLOADS]: The skill interacts with the GitHub API to fetch and update issue content. GitHub is a well-known and trusted service, and this communication is essential for the skill's stated purpose.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting and acting upon data retrieved from external GitHub issues which could be controlled by third parties.
Ingestion points: The agent reads the titles, bodies, and comments of GitHub issues (the "map" and "tickets") from a repository. These fields can contain arbitrary text.
Boundary markers: There are no explicit boundary markers or instructions to the agent to treat the ingested issue content as untrusted or to ignore any embedded instructions within that content.
Capability inventory: The skill possesses the ability to execute shell commands via the gh CLI and is specifically instructed to "invoke the skills the ## Notes block names," where those names are dynamically sourced from the issue body.
Sanitization: No evidence of sanitization, escaping, or validation of the content retrieved from GitHub issues before it is used to guide the agent's behavior or decision-making.

wayfinder