zynkr-kms

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from Gmail threads to generate knowledge base entries. While this presents an injection surface, the risk is mitigated by a mandatory manual review process where the user must approve all proposed changes before they are committed to the knowledge base.
  • Ingestion points: Gmail thread content retrieved via mcp__google-workspace__get_gmail_thread_content in SKILL.md (Step 2).
  • Boundary markers: None explicitly defined in prompts, but manual user approval acts as a logical gate.
  • Capability inventory: File system writes (metadata updates), network operations (KB updates via zynkr MCP), and Gmail label modification via google-workspace MCP.
  • Sanitization: Implements automated PII stripping for emails, phone numbers, and attendee lists before drafting content.
  • [DATA_EXFILTRATION]: The skill accesses sensitive Gmail communications and transfers processed content to the Zynkr platform. This behavior is consistent with the skill's primary purpose, and all external resources (the Zynkr platform and the deployment repository) are owned by the skill's author.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 07:31 PM
Security Audit — agent-trust-hub — zynkr-kms