wayfinder
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from an external issue ledger, including ticket titles, question bodies, and mapping notes, which are then used to guide agent decisions.
- Ingestion points: Issue ledger content, specifically ticket titles and bodies processed during mapping (SKILL.md).
- Boundary markers: The instructions lack defined delimiters or safety directives to prevent the agent from obeying instructions embedded within the issue content.
- Capability inventory: The skill can execute local scripts via
frontier.pyand trigger subagents such as/researchor/agent-reachbased on the ledger state (SKILL.md). - Sanitization: No sanitization, escaping, or validation of the ingested ticket content is performed before interpolation into the agent's context.
- [COMMAND_EXECUTION]: The skill documentation references the execution of a local script,
frontier.py, to identify the next priority ticket in the mapping process. This indicates a functional dependency on local executable code within the agent's environment.
Audit Metadata