governance-encoder
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes human-readable governance documentation from potentially untrusted or peer-editable sources to generate behavioral constraints.
- Ingestion points: The skill reads
CONTRIBUTING.md, wiki pages, PR templates, issue templates, and Figma library rules to extract governance patterns. - Capability inventory: The skill has the capability to write files to the
.ai/governance/directory, which are then used by the agent and CI pipelines to enforce project constraints. - Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are defined for the content extracted from external documentation files.
- Sanitization: There is no evidence of sanitization or filtering of the content pulled from documentation sources before it is interpolated into the rule-generation logic. A malicious actor could potentially insert instructions into a project's wiki or PR templates designed to manipulate the resulting rule engine's severity or logic.
Audit Metadata