maestro-help
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill operates as a read-only information provider. It utilizes file-access tools (Read, Grep, Glob) to retrieve documentation from its own package and project metadata from the
.workflow/directory. It does not have access to shell execution or network tools.\n- [INDIRECT_PROMPT_INJECTION]: The skill analyzes.workflow/state.jsonto offer relevant command suggestions. While this involves processing local user data, the skill's lack of executable capabilities prevents any significant risk of prompt injection beyond informational bias.\n - Ingestion points: Reads
.workflow/state.jsoninphases/01-parse-intent.md.\n - Boundary markers: None explicitly defined in the skill instructions.\n
- Capability inventory: Limited to Read, Grep, and Glob for information retrieval.\n
- Sanitization: Inputs are restricted to internal state fields and keyword matching against the command catalog.
Audit Metadata