framer-deploy-integration

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses platform-specific CLI tools to manage deployments and configuration.
  • Evidence: Use of vercel env add, vercel --prod, fly secrets set, and fly deploy to interact with cloud environments.
  • The allowed-tools field correctly scopes these commands to the vercel, fly, and gcloud namespaces.
  • [INDIRECT_PROMPT_INJECTION]: The provided code templates for Vercel functions and webhook handlers ingest external data which could contain malicious instructions.
  • Ingestion points: The skill provides logic to process req.body in api/sync-framer.ts and api/webhook-handler.ts.
  • Boundary markers: Absent; the templates do not include delimiters or instructions for the agent to ignore content within the ingested data.
  • Capability inventory: The skill has the ability to write files, edit configuration, and execute platform CLI commands (Bash).
  • Sanitization: No input validation or sanitization is implemented in the simplified example code, relying on the user to add these protections.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 03:46 AM
Security Audit — agent-trust-hub — framer-deploy-integration