birdview-review
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a local JavaScript file located at
../birdview/scripts/render-review.mjs. This script is used to generate and display the architecture review findings. As this refers to a local file within the expected directory structure of the vendor's toolset, it is considered standard functionality. - [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it analyzes external project source code to generate diagrams and recommendations.
- Ingestion points: Project source code and file boundaries analyzed during the review process.
- Boundary markers: The instructions do not define specific delimiters for project content, but they do require the agent to wait for user confirmation before implementing any proposed changes.
- Capability inventory: The skill can read local files and execute specific local scripts (
render-review.mjs). - Sanitization: No explicit sanitization or filtering of project content is mentioned in this entry file.
Audit Metadata