affinity-api
Pass
Audited by Gen Agent Trust Hub on Mar 12, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core functionality of processing external CRM data.
- Ingestion points: The skill facilitates reading untrusted content from several sources including CRM notes (
references/v2-notes.md), emails and chat messages (references/v2-interactions.md), and meeting transcripts (references/v2-transcripts.md). - Boundary markers: There are no instructions or delimiters provided in the documentation to help the agent distinguish between system instructions and data retrieved from the CRM.
- Capability inventory: The agent has the capability to read and write sensitive business data (persons, organizations, opportunities) and manage files via the
api.affinity.coendpoints. - Sanitization: No sanitization or validation logic is defined for the incoming data streams.
- [NO_CODE]: The skill does not contain any executable scripts, binaries, or configuration files that trigger code execution. It serves as a knowledge base for API interactions.
Audit Metadata