The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (conversation context, codebase content, and docs/agents/issue-tracker.md) to generate a specification. While it uses a structured template, a malicious codebase or document could attempt to influence the agent's behavior during the synthesis process.
Ingestion points: Conversation context, docs/agents/issue-tracker.md, and repo files.
Boundary markers: The skill uses a defined <spec-template> which helps structure the output.
Capability inventory: Shell command execution via gh issue create and glab issue create.
Sanitization: Not explicitly defined in instructions, relying on the platform's standard CLI tool handling.
[COMMAND_EXECUTION]: The skill uses gh (GitHub CLI) and glab (GitLab CLI) to create issues. These are standard developer tools and used for the skill's primary stated purpose, though they provide a mechanism for the agent to interact with external services.

to-spec