open-code-review-delegate
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external source code and diffs, which represents a surface for indirect prompt injection. A malicious actor could embed instructions in code comments to influence the agent's review process.
- Ingestion points: Reads file content via
catandgit diffcommands, and rules via theocrCLI (SKILL.md). - Boundary markers: The skill does not define specific delimiters or instructions for the agent to ignore content within the ingested code.
- Capability inventory: The agent has access to shell commands including
git,cat, and theocrCLI (SKILL.md). - Sanitization: The instructions include specific guidance for the agent to use shell-safe arguments and avoid double-quoted templates when handling background context to prevent command injection (SKILL.md, "Recovering Oversized Background Context").- [EXTERNAL_DOWNLOADS]: Recommends the installation of the
@alibaba-group/open-code-reviewpackage via NPM. - Evidence:
npm install -g @alibaba-group/open-code-review(SKILL.md). - Context: This is the official CLI tool for the skill, hosted on a well-known registry and provided by the skill's author.
Audit Metadata