lean-canvas
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's 'UPDATE' and 'FIND' modes involve reading existing markdown files from the filesystem. This creates a surface where malicious instructions embedded in a canvas file could influence agent behavior.
- Ingestion points: The agent reads file content when a path is provided (e.g.,
@path/to/lean-canvas.md) or when searching theproduct/,strategy/,canvas/, andplanning/directories. - Boundary markers: The skill lacks explicit instructions or delimiters to isolate data read from files, such as 'treat the following file content as data only'.
- Capability inventory: The skill possesses the ability to search directories, read file contents, and write/save markdown files to the local system.
- Sanitization: No sanitization or validation logic is defined for the content retrieved from existing files before it is processed by the agent.
Audit Metadata