The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external Product Requirements Documents (PRDs) stored in GitHub issues or sidecar work items. These external sources are untrusted and can be modified by third parties. The skill does not provide the agent with explicit instructions to use boundary markers or sanitization when processing this data, creating a surface for indirect prompt injection where instructions embedded in the PRD could influence the agent's actions.
Ingestion points: The skill fetches PRD content from external sources in SKILL.md using gh issue view <number> for GitHub-based state and sidecar.sh work view "$PRD" --json for sidecar-based state.
Boundary markers: The instructions do not mandate the use of delimiters (e.g., XML tags or triple quotes) or provide warnings to the agent to ignore instructions contained within the ingested text.
Capability inventory: The skill provides the agent with capabilities to modify the environment, including creating and editing GitHub issues (gh issue create, gh issue edit) and performing various state-modifying operations on the local sidecar repository via the sidecar.sh script.
Sanitization: The skill lacks mechanisms for filtering or sanitizing external content before it is interpolated into the agent's context.

prd-to-issues