The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external, potentially untrusted data to perform its core functions without explicit isolation or sanitization.
Ingestion points: The skill ingests data from git diff outputs, conversation history, issue trackers, specifications, and project plans (referenced in the 'Review' and 'Check requirement coverage' sections).
Boundary markers: There are no instructions to use specific delimiters (like XML tags or markdown blocks) or explicit warnings to disregard instructions embedded within the ingested data.
Capability inventory: The skill uses gh pr create for network-based CLI interaction and standard file system read operations through git tools.
Sanitization: There is no explicit sanitization or filtering of the input data before it is used to influence the agent's logic or included in the generated pull request body.

dev-ship