defend-it
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use shell commands, specifically
git diff, to analyze source code changes and surrounding context. These are standard development operations and are necessary for the skill's stated purpose. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (pull requests, git diffs, and documents) which represents a potential surface for indirect prompt injection.
- Ingestion points: Analyzes artifact content including PR bodies, git diffs, and document paths (SKILL.md).
- Boundary markers: Absent; there are no explicit instructions to treat the ingested artifact content with specific delimiters or as untrusted text.
- Capability inventory: The skill has the capability to read and run code and execute shell commands to build a 'key' for the defense process.
- Sanitization: Absent; the skill directly processes the artifact content to generate quiz items and verify claims.
Audit Metadata