infrahub-teaching-concepts
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is authored by OpsMill and correctly targets their own product ecosystem (Infrahub) using official documentation (docs.infrahub.app) and CLI tools (infrahubctl).
- [COMMAND_EXECUTION]: The skill utilizes
infrahubctlvia theBashtool for verifying learner exercises and interacting with Infrahub instances. These actions are protected by critical safety rules (inrules/safety-instance-writes.md) that mandate explicit user consent, the use of isolatedlearning-*branches, and a mandatory cleanup step to delete scratch branches. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the user's repository (schema YAML, object files, and
.infrahub.yml) to generate personalized lessons and exercises. While this establishes an ingestion surface for potentially malicious instructions, the risk is mitigated by the skill's rigid operational protocol (Probe/Explain/Exercise/Check sequence) and safety constraints that prevent data-driven execution on the default branch or sensitive environments.
Audit Metadata