The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the GitHub CLI (gh) to automate the creation and management of tracking issues.
Evidence includes gh issue create, gh label create, gh issue edit, and gh issue comment commands used to synchronize review state with the repository.
The use of quoted heredocs (cat <<'EOF') in shell commands demonstrates a secure practice by preventing shell variable expansion and command substitution within the issue body.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data in the form of code review findings, line numbers, and code snippets.
Ingestion points: Finding details and evidence snippets are ingested in Step 2 and Step 3 to populate the tracking issue template.
Boundary markers: The skill uses markdown tables and triple-backtick code blocks to delimit user-provided content from the structural template.
Capability inventory: The agent has the ability to write to the repository's issue tracker and labels via the gh tool.
Sanitization: The instructions recommend using quoted heredocs for CLI operations, which mitigates simple command injection during shell execution.
[SAFE]: All identified behaviors are consistent with the skill's primary purpose of issue-driven development and technical debt tracking. No malicious patterns, obfuscation, or unauthorized data access were detected.

deferred-finding