The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill reads and processes external data that could be controlled by an attacker, such as git commit messages, GitHub issue descriptions, and comments. This data is used to determine the scope of work and recommend next actions.
Ingestion points: The skill uses git log, git status, git diff, gh issue view, gh pr list, and gh api to gather situational awareness. It also reads project documentation files like AGENTS.md and README.md in the current repository.
Boundary markers: There are no explicit boundary markers or instructions to ignore embedded commands within the ingested data defined in the skill instructions.
Capability inventory: The skill has the capability to read files, run git commands, and use the GitHub CLI. While it primarily scouts for information, it can transition to implementation tasks (file writes) and GitHub interactions (comments, branch management) if authorized by the user.
Sanitization: No explicit sanitization or validation of the external content is performed before it is processed by the agent to define the plan.

recon