wayfinder
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from an external issue tracker, creating a surface for indirect prompt injection.
- Ingestion points: The agent reads the 'map' issue and individual 'decision ticket' bodies from the repository's issue tracker to orient itself and resolve tasks.
- Capability inventory: The agent has the ability to create and modify issues, load and apply other skills (grill, research, prototype, domain-modeling), and potentially execute shell commands via the 'task' ticket type or the setup command.
- Boundary markers: The skill lacks explicit boundary markers or instructions telling the agent to treat issue content purely as data and to ignore any embedded natural language instructions.
- Sanitization: There is no evidence of sanitization, filtering, or validation of the content retrieved from the issue tracker before it is processed by the agent.
- [COMMAND_EXECUTION]: The skill references the execution of
/setup-snappedly-skills. While this is a vendor-provided tool for environment configuration, it represents a shell command execution pattern during the setup phase.
Audit Metadata