The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests untrusted external data to drive its ticket generation and publishing logic.
Ingestion points: Processes user-provided references, specification paths, issue numbers, and external URLs (documented in SKILL.md, Step 1).
Capability inventory: Writing local markdown files to .scratch/ and interacting with issue tracker APIs (GitHub, Linear) to create issues (documented in SKILL.md, Step 5).
Boundary markers: The instructions do not define specific delimiters or instructions to ignore potential commands embedded within the external context.
Sanitization: No evidence of sanitization or validation of the fetched external content before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill facilitates fetching content from external sources provided by the user.
Evidence: SKILL.md instructs the agent to fetch and read the full body and comments of URLs or issue references passed as arguments.
[SAFE]: The skill demonstrates professional development practices, including a detailed provenance record and clear versioning.
Evidence: The presence of PROVENANCE.md, UPSTREAM.json, and ADAPTATIONS.patch indicates a transparent supply chain and controlled modification process based on guidance from OpenAI.

to-tickets