managing-integrations
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the processing and modification of integration metadata, scripts, and configurations that may originate from untrusted sources, creating a surface for indirect instructions to influence agent behavior.
- Ingestion points: The agent reads data from the platform using
celigo integrations list/getand from local files usingceligo pull, which includes integration JSON, JavaScript scripts, and Markdown prompts. - Boundary markers: The instructions do not specify delimiters or warnings for the agent to ignore instructions embedded within the processed data or scripts.
- Capability inventory: The skill provides extensive write capabilities, including
celigo integrations update,celigo push, andceligo promote, allowing the agent to modify environment configurations and resource logic across multiple environments. - Sanitization: There is no documented process for validating or escaping content retrieved from integrations or local files before it is used in subsequent operations or pushed back to the Celigo environment.
Audit Metadata