design-council-orchestration
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill installation instructions refer to a third-party GitHub repository for cloning and plugin installation (github.com/sjsyrek/design-council).
- [COMMAND_EXECUTION]: The skill executes shell commands using the beads CLI tool (such as 'bd memories' and 'bd create') to synchronize project state and manage task tracking when the tool is present on the user's system.
- [INDIRECT_PROMPT_INJECTION]: The skill aggregates information from various project files to inform its background agents, creating a potential surface for indirect instruction ingestion. 1. Ingestion points: The skill reads CLAUDE.md, project specifications, and automated memory into a brief artifact. 2. Boundary markers: No explicit delimiters or isolation instructions are mentioned for the ingested content. 3. Capability inventory: The skill can spawn background agents, write to the file system at ~/.claude/councils/, and execute CLI commands. 4. Sanitization: No specific filtering or escaping logic for external data is described.
- [DATA_EXPOSURE]: The skill maintains logs and briefs within the local file system (~/.claude/councils/) to coordinate between multiple agent contexts.
Audit Metadata