context-canary

Pass

Audited by Gen Agent Trust Hub on Jun 12, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely instructional, establishing a 'canary contract' where the agent prefixes responses with status markers (name, turn counter, self-check) to signal context health and instruction adherence.
  • [SAFE]: External references in the research document point to legitimate research papers, established technology companies, and well-known developer platforms for educational context.
  • [SAFE]: The 'trip protocol' recommends standard project management practices, such as writing state checkpoints to local files like HANDOFF.md, which does not involve unauthorized data access or exfiltration.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 12, 2026, 08:20 PM
Security Audit — agent-trust-hub — context-canary