skills/danielmeppiel/genesis/genesis/Gen Agent Trust Hub

genesis

Pass

Audited by Gen Agent Trust Hub on Sep 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to process user-supplied 'operator prompts,' which creates a potential surface for indirect prompt injection. This is addressed by a mandatory 8-step design discipline that utilizes deterministic validation gates (S4), attention anchors (B8), and tool-call bridges (S7) to ensure that all side-effecting actions are governed by strict schemas rather than free-form model output. The ingestion point is restricted to Step 1, with findings persisted in a structured handoff packet that serves as a boundary between reasoning and execution.\n- [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of dependencies and adapters from external sources. It specifically references the 'apm-guide' from the 'microsoft/apm' repository, which belongs to a trusted organization. It also relies on well-known services such as 'agentskills.io' for canonical specifications and 'vals.ai' for model performance grounding. These external references follow best practices, such as version pinning (e.g., eslint@9.0.0), and are documented in an informative, non-alarming manner.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 28, 2026, 05:28 AM