LaunchDarkly Automation
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external LaunchDarkly APIs which could potentially contain malicious instructions.
- Ingestion points: Data is ingested through tools like
LAUNCH_DARKLY_LIST_PROJECTS,LAUNCH_DARKLY_GET_ENVIRONMENTS, andLAUNCH_DARKLY_LIST_CODE_REFERENCE_REPOSITORIESwhich fetch project names, environment slugs, and code references. - Capability inventory: The skill can perform state-changing operations such as creating (
LAUNCH_DARKLY_CREATE_TRIGGER_WORKFLOW) and deleting (LAUNCH_DARKLY_DELETE_TRIGGER_WORKFLOW) flag triggers. - Boundary markers: The instructions do not define explicit delimiters or warnings to separate data retrieved from APIs from the agent's instructions.
- Sanitization: There is no mention of sanitization or validation of the data retrieved from external sources before it is processed by the agent.
- [SAFE]: The skill uses official vendor domains (
composio.dev) and services (rube.app) for its operations. - [SAFE]: Authentication is handled correctly by prompting the user for an API key rather than hardcoding credentials or using unsafe storage patterns.
Audit Metadata