swarm-planner
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from multiple sources to generate implementation plans and native tasks. This creates a surface for indirect prompt injection attacks.
- Ingestion points: The skill uses
Explore,Glob, andGrepto read local codebase files, andWebSearchto retrieve external documentation from arbitrary websites. - Boundary markers: The instructions do not specify any delimiters or safety warnings to help the model distinguish between implementation requirements and potentially malicious instructions embedded in the analyzed code or documentation.
- Capability inventory: The skill has the capability to create and modify native agent tasks (
TaskCreate,TaskUpdate), spawn subagents (Task), and write files to the local system (docs/plans/). - Sanitization: There is no explicit sanitization or filtering of the content retrieved from external sources before it is incorporated into the task descriptions or validation commands.
Audit Metadata