The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Python script located at .claude/skills/gh/scripts/github_project_setup.py using uv run. This script is invoked to perform actions such as closing milestones, reassigning issues, and updating project statuses.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources (GitHub milestones and issues) that could contain malicious instructions designed to influence the agent's behavior.
Ingestion points: Untrusted data enters the context through backlog_list_milestones and backlog_list_issues as described in Step 1 of the workflow.
Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded content in the fetched data.
Capability inventory: The skill possesses the capability to execute shell commands via the github_project_setup.py script and modify GitHub resources (milestones, issues, and Project V2 items).
Sanitization: There is no explicit logic described to sanitize, filter, or escape the data retrieved from GitHub before it is interpolated into summaries or script arguments.

complete-milestone