iblai-api-agent-prompt
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFEDATA_EXFILTRATIONCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill uses curl commands to interact with the vendor's official domain (api.iblai.app). This is the intended behavior for agent management and does not involve unauthorized data movement.
- [CREDENTIALS_UNSAFE]: The skill correctly uses environment variables such as $IBLAI_API_KEY and $IBLAI_ORG for authentication, avoiding the use of hardcoded secrets.
- [INDIRECT_PROMPT_INJECTION]: As the skill is designed to update agent prompts and instructions, it theoretically handles untrusted data that could contain prompt injections. However, this is part of its core functionality for prompt management.
- Ingestion points: system_prompt and suggested prompt fields in SKILL.md.
- Boundary markers: None specified in the documentation.
- Capability inventory: Platform API interactions via PUT and POST.
- Sanitization: Relies on backend API validation.
Audit Metadata