confluence-expert

Pass

Audited by Gen Agent Trust Hub on Apr 20, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides legitimate instructions and code examples for interacting with official Atlassian APIs. No malicious patterns or bypass techniques were detected.
  • [CREDENTIALS_UNSAFE]: Documentation regarding authentication methods (API tokens, PATs, and OAuth 2.0) uses clear placeholders and explicitly recommends against hardcoding secrets, advocating for secure management via environment variables.
  • [EXTERNAL_DOWNLOADS]: All external references and dependencies target trusted Atlassian domains and official package registries. Packages mentioned include 'atlassian-python-api', 'requests', and '@forge/api'.
  • [PROMPT_INJECTION]: The skill facilitates data ingestion from Confluence content; however, this is inherent to its documented purpose and no instructions to override agent behavior were found.
  • Ingestion points: Confluence REST API responses (pages, comments, blogposts) and webhook payloads.
  • Boundary markers: Not explicitly defined in example code.
  • Capability inventory: CRUD operations on Confluence pages and spaces, permission management, and webhook registration.
  • Sanitization: Not present in the provided reference snippets.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 20, 2026, 11:27 AM
Security Audit — agent-trust-hub — confluence-expert