swarm-advanced
Pass
Audited by Gen Agent Trust Hub on Mar 1, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill recommends a global installation of the
claude-flowpackage from the npm registry, which is the foundational tool for the orchestration patterns described.\n- [COMMAND_EXECUTION]: The orchestration logic relies on tools likemcp__claude-flow__parallel_executeandmcp__claude-flow__workflow_executethat trigger command-line operations and task sequences across multiple agents.\n- [PROMPT_INJECTION]: The research and analysis patterns create a surface for indirect prompt injection. Agents are tasked with gathering data from external sources like web publications and academic papers, which could contain adversarial instructions.\n - Ingestion points: External data is collected via
mcp__claude-flow__parallel_executeand stored in memory usingmcp__claude-flow__memory_usage.\n - Boundary markers: The examples do not demonstrate the use of delimiters or system instructions to ignore potential commands within the collected data.\n
- Capability inventory: The framework allows for agent spawning, task orchestration, file system interactions via
mcp__claude-flow__memory_backup, and neural network training.\n - Sanitization: There is no explicit sanitization or validation of the data retrieved from external sources before it is processed by the analysis agents.
Audit Metadata