release-notes-generator
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data such as commit messages, work item descriptions, and source code diffs, which could contain instructions designed to influence the agent's behavior.
- Ingestion points: Commit titles from
dxs source compare, work item titles and descriptions fromdxs source workitems, and source code content fromdxs source diff. - Boundary markers: There are no explicit instructions or delimiters defined to isolate the untrusted external content from the agent's operational instructions.
- Capability inventory: The skill uses
dxsCLI tools to read repository metadata and code, and generates formatted text as output. - Sanitization: No validation or filtering mechanisms are described for the data fetched from Azure DevOps or the repository before it is processed by the agent.
Audit Metadata