devin-review-init
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to Indirect Prompt Injection. It is designed to ingest data from untrusted sources within the repository to inform its generation logic.
- Ingestion points: The skill reads
REVIEW.md,AGENTS.md,CLAUDE.md,CONTRIBUTING.md,.cursorrules,.windsurfrules, and other*.rulesor*.mdcfiles from the repository being analyzed. - Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are defined for the agent to use when reading these files.
- Capability inventory: The skill uses the agent's underlying capabilities to read files and write new configuration files (
REVIEW.md,AGENTS.md). - Sanitization: There is no evidence of sanitization or filtering of the content ingested from the existing repository files.
- [COMMAND_EXECUTION]: The exploration phase of the skill (Phase 1) suggests the use of command-line tools like
ls -laand searching through directory structures (docs/,apps/, etc.) to map the repository. While these are read-only exploration commands, they are part of the skill's automated workflow.
Audit Metadata