runway-common-errors
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements best practices for handling sensitive data. It explicitly instructs the agent to 'redact bearer secrets, unsafe prompt bodies, customer media, and signed output URLs' during incident triage and output generation.
- [SAFE]: Tool discipline is strictly defined. The skill limits the use of
WriteorEdittools to approved implementations and explicitly forbids mutating production Runway resources (creating, deleting, or retrying tasks) without operator approval. - [SAFE]: External references target official documentation at
docs.dev.runwayml.com. These are recognized as well-known service domains and do not escalate security risk. - [SAFE]: The skill emphasizes operational safety by requiring reconciliation of durable task IDs before any retry attempt to prevent accidental duplicate generation and credit waste.
Audit Metadata