The Agent Skills Directory

[SAFE]: The skill utilizes the GitHub CLI (gh) to interact with GitHub, which is a well-known and trusted service. This behavior is expected and aligns with the skill's documented purpose of publishing project specifications.
[PROMPT_INJECTION]: The skill presents a potential surface for indirect prompt injection as it processes data from untrusted sources (conversation history and codebase).
Ingestion points: The skill reads the current conversation context and explores the repository files.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to distinguish between its primary instructions and the ingested data.
Capability inventory: The agent has permissions to explore the repository and publish content using the gh tool.
Sanitization: The instructions do not include specific steps to sanitize or filter the synthesized content before it is published to the issue tracker.

to-issue