The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data in the form of pull request diffs and source code files.
Ingestion points: Untrusted content enters the agent's context through gh pr diff, gh pr view, and direct file reads during code review.
Boundary markers: Absent. The instructions do not specify the use of delimiters or provide explicit warnings to the agent to ignore instructions embedded within the reviewed code.
Capability inventory: The skill possesses write capabilities including posting reviews via the GitHub API (gh api) and applying local code fixes via the fix command in the interactive walkthrough.
Sanitization: Absent. The skill does not explicitly sanitize or validate the content of the reviewed files before processing.
[DYNAMIC_EXECUTION]: The skill includes a shell script that performs runtime generation of an HTML report.
Evidence: The generate-report.sh script executes a Python one-liner to inject JSON data into a template: python3 -c "..." "$TEMPLATE" "$JSON_FILE" "$OUTPUT".
Context: This behavior is limited to local report generation from established templates and does not involve remote code execution or execution of untrusted logic.

panel-review