The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from GitHub pull request comments and review threads and uses it to guide its actions without adequate security boundaries.
Ingestion points: The skill retrieves PR data, inline comments, and issue comments using gh pr view, gh api, and GraphQL queries in SKILL.md.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to disregard or treat carefully any commands or instructions found within the ingested comments.
Capability inventory: The skill has access to the Bash, Edit, and Write tools. It performs sensitive operations including gh pr checkout, git commit, git push, and gh api -X DELETE in SKILL.md.
Sanitization: The skill does not implement any validation, filtering, or escaping of the content retrieved from external GitHub comments before the agent processes it.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute GitHub CLI (gh) and git commands. While these are necessary for the skill's primary function of PR management, they represent a significant capability that could be abused if the agent is influenced by malicious input in PR comments.
Evidence: Execution of gh pr view, gh pr checks, gh api, gh pr checkout, git commit, git push, gh pr edit, gh pr comment, and gh api -X DELETE in SKILL.md.

my-pr-checker