metronome-plg-billing
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill acts as a surface for indirect prompt injection by ingesting untrusted external data during the pricing discovery phase.
- Ingestion points: The agent is instructed to visit user-provided pricing page URLs and parse prose descriptions in SKILL.md (Phase 1).
- Boundary markers: The skill implements a manual confirmation gate, requiring the founder to verify the agent's interpretation of the pricing model before any configuration occurs.
- Capability inventory: The agent has wide-ranging read/write capabilities via the Metronome API, including the creation of billable metrics, rate cards, and contracts as detailed in worked-examples.md.
- Sanitization: No explicit sanitization, filtering, or escaping mechanisms for the ingested external data are described beyond the user confirmation step.
Audit Metadata