consult-launch-comms

Pass

Audited by Gen Agent Trust Hub on Sep 10, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from Google Docs in the client project folder (e.g., [UAT], [Plan], [BRD]) to draft communications. 1. Ingestion points: mcp__google-workspace__get_doc_content reads files from the Drive project folder. 2. Boundary markers: Absent. 3. Capability inventory: Doc creation via mcp__google-workspace and SQL updates via mcp__supabase. 4. Sanitization: A mandatory human-review gate is enforced at Step 4 before any drafts are finalized.
  • [COMMAND_EXECUTION]: The skill updates CRM deal notes using SQL queries via mcp__supabase. It includes explicit developer instructions to escape single quotes in project data to prevent SQL injection.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: Hardcoded identifiers for Supabase projects and Google Drive folder IDs are used for internal workflow routing within the author's ecosystem.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 10, 2026, 07:31 PM
Security Audit — agent-trust-hub — consult-launch-comms