anth-webhooks-events

Pass

Audited by Gen Agent Trust Hub on May 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill functions as a technical guide and provides boilerplate code for integrating with Anthropic's API events.
  • [EXTERNAL_DOWNLOADS]: The skill references official Anthropic documentation and examples using standard, reputable Python libraries (anthropic, redis, rq, requests). All references are consistent with the skill's educational purpose.
  • [DATA_EXFILTRATION]: Includes a boilerplate example using the requests library to send data to an internal callback URL. This is a standard architectural pattern for event-driven systems and does not constitute unauthorized exfiltration.
  • [PROMPT_INJECTION]: Indirect prompt injection surface analysis:
  • Ingestion points: The provided code examples demonstrate processing variables (prompt, documents) which could originate from external sources.
  • Boundary markers: Not included in the abstract code snippets.
  • Capability inventory: Examples use the anthropic SDK and requests library.
  • Sanitization: No specific sanitization or filtering logic is included in the educational boilerplate.
Audit Metadata
Risk Level
SAFE
Analyzed
May 19, 2026, 11:07 PM
Security Audit — agent-trust-hub — anth-webhooks-events