Swarm Orchestration for Linear
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external issue trackers (Linear), which constitutes an indirect prompt injection surface as issue titles and descriptions are attacker-controlled content.
- Ingestion points: The skill ingests issue titles and keys from a Linear backlog (SKILL.md).
- Boundary markers: The architecture utilizes git worktrees for filesystem isolation and recommends explicit identifier validation.
- Capability inventory: The skill utilizes the Bash tool for repository management and git operations.
- Sanitization: The instructions explicitly advise developers to reject malformed identifiers and treat issue titles as untrusted input rather than attempting to sanitize them.
- [EXTERNAL_DOWNLOADS]: The skill references the OpenAI Symphony repository on GitHub as an architectural pattern origin. This is a reference to a well-known service for design patterns.
Audit Metadata