generate-integration-docs

Pass

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it requires the agent to read and summarize content from potentially untrusted external sources (internet search results) and a variety of local files.
  • Ingestion points: The skill explicitly instructs the agent to search the internet for API versioning and deprecation schedules (SKILL.md). It also reads configuration and source code from sibling repositories including rudder-integrations-config, rudder-integrations-info, and rudder-auth (SKILL.md, references/integration-docs-guide.md).
  • Boundary markers: There are no protective delimiters or instructions to ignore embedded commands within the ingested data, which could allow a malicious file or website to influence the agent's output or behavior.
  • Capability inventory: The agent has the capability to write to the local filesystem to generate the requested documentation files (README.md, businesslogic.md, retl.md).
  • Sanitization: The skill lacks mechanisms to sanitize or validate the content retrieved from external searches before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 21, 2026, 03:10 AM
Security Audit — agent-trust-hub — generate-integration-docs