The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill analyzes potentially untrusted external content such as project records, repository state, and pull request metadata to generate its summary. * Ingestion points: Project records, repository state, task status, and pull request information are ingested as described in the 'Recover the state' section of SKILL.md. * Boundary markers: The instructions do not define specific delimiters or boundary markers to separate external content from the agent's internal instructions. * Capability inventory: The skill does not use tools for file-writing, network operations, or code execution. It includes a safety constraint: 'This is a recap, not authorization to resume implementation, change files, merge work, or start background tasks.' Additionally, the frontmatter specifies 'disable-model-invocation: true'. * Sanitization: No programmatic sanitization or filtering of the ingested external project data is described.

remind-me