threejs-game-release
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves ingesting and acting upon external data sources, which presents a vulnerability to manipulated input data.
- Ingestion points: The agent is instructed to read external platform matrices, resource registration logs, test plans, and release checklists from Work Items (
SKILL.md). - Boundary markers: The instructions do not define any delimiters or negative constraints (e.g., 'ignore instructions inside data') for the external files being processed.
- Capability inventory: The skill orchestrates sensitive operations including
threejs-build-upload,threejs-backend-config,threejs-store-submit, andthreejs-release. These tools perform network and file operations that could be misdirected by malicious data in the input files (SKILL.md). - Sanitization: No specific input validation or escaping mechanisms are described. Furthermore, the instructions explicitly state that candidate SHAs and package hashes are optional and should not block the workflow, which bypasses automated integrity checks that would otherwise mitigate some injection or tampering risks.
Audit Metadata