crm-adapter-airtable

Pass

Audited by Gen Agent Trust Hub on Apr 29, 2026

Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes data from external sources (Airtable) and local files, which presents a surface for indirect prompt injection if the synchronized content contains malicious instructions.
  • Ingestion points: Data is read from crm/companies.md, crm/opportunities.md, crm/outreach.md, and Airtable API responses.
  • Boundary markers: Absent. The skill instructions do not define delimiters or specific instructions to treat record values as untrusted data.
  • Capability inventory: The skill uses the Write tool to modify local markdown files and the .airtable-sync.json manifest, and utilizes an Airtable MCP/CLI for remote data operations.
  • Sanitization: Absent. There is no logic specified for validating or sanitizing the content of the records before processing.
  • [DATA_EXFILTRATION]: Transfers CRM data from local markdown files to an external Airtable base as part of its primary synchronization function. This requires the configuration of an Airtable Base ID and credentials.
  • [COMMAND_EXECUTION]: Uses the Bash tool restricted to test and date commands to verify file existence and generate timestamps for synchronization metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 29, 2026, 05:16 AM
Security Audit — agent-trust-hub — crm-adapter-airtable