wayfinder
Pass
Audited by Gen Agent Trust Hub on Aug 18, 2026
Risk Level: SAFEPROMPT_INJECTIONCREDENTIALS_UNSAFENO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external work trackers, which acts as a surface for indirect prompt injection if the tracker contains malicious instructions.
- Ingestion points: Tracker records and ticket bodies referenced in SKILL.md.
- Boundary markers: Absent; there are no delimiters to isolate external text from agent instructions.
- Capability inventory: The agent can update tracker records, spawn research subagents, and execute prototypes.
- Sanitization: Absent; no validation is performed on the ingested tracker data.
- [DATA_EXPOSURE_AND_EXFILTRATION]: The resolution process for Task tickets includes recording the locations of credentials. While intended for coordination, documenting where secrets are stored in a shared log increases the risk of unauthorized access.
Audit Metadata