swarm-planner
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from the local codebase and external documentation to inform its planning process.
- Ingestion points: The skill reads from the project codebase (Research step) and external libraries/APIs via Context7, MCP, or web search (Documentation step).
- Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from following malicious instructions that might be embedded in the codebase files or fetched documentation.
- Capability inventory: The skill has the ability to write files to the local directory (Save Plan step) and spawn subagents to process the plan (Subagent Review step).
- Sanitization: The instructions do not specify any validation, sanitization, or filtering for content retrieved from external sources or the codebase before it is used in the plan generation or review phase.
Audit Metadata