The Agent Skills Directory

[SAFE]: The skill provides procedural guidance for software development and does not include any executable code, shell scripts, or direct network operations. It encourages best practices for secret management by advising users not to commit living specification documents to the repository.- [INDIRECT_PROMPT_INJECTION]: The workflow involves summarizing user conversations into external issue trackers, which represents a potential ingestion surface for untrusted data. 1. Ingestion points: User-provided "grilled" conversations (SKILL.md). 2. Boundary markers: No explicit delimiters or instructions to ignore embedded commands are specified in the workflow. 3. Capability inventory: Writing documentation and tickets to external systems like GitHub (via gh CLI), Linear, or Jira. 4. Sanitization: No specific content validation or escaping steps are defined in the manual process.

to-spec