sentinel-change-map
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes repository source code and diffs to generate visualization maps, creating an interface where data from the codebase could influence the agent's summary.
- Ingestion points: Repository source code,
state.json, and local.sentinelfiles. - Boundary markers: The
rendercommand uses a--strictflag to validate inputs and flow consistency. - Capability inventory: The skill executes
sentinelCLI commands and writes to the local filesystem. - Sanitization: The workflow includes a specific check for sensitive files (e.g.,
.env, keys, credentials) in thecollectstep and refuses to process them if found. - [COMMAND_EXECUTION]: The skill utilizes the
sentinelcommand-line tool for data collection, diff inspection, and HTML rendering. These actions are restricted to the local workspace and intended for development visualization.
Audit Metadata