moshu-volume
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes local Python utility scripts (
design_fingerprints.py,impact_scan.py,pace_meter.py,tracking_commit.py) to automate writing project maintenance. These scripts perform safe, deterministic operations like calculating SHA-256 hashes of Markdown files to detect setting changes, scanning text for narrative consistency, and managing a JSON-based tracking state. All file operations are confined to the project directory. - [INDIRECT_PROMPT_INJECTION]: The skill includes a research module ('采风') that gathers creative inspiration from external sources via a sub-agent. While this involves ingesting untrusted data, the skill mandates strict 'translation' protocols and anti-plagiarism checks to sanitize and re-contextualize the information before it is presented to the user or used in the draft, effectively mitigating injection risks.
- [REMOTE_CODE_EXECUTION]: The skill does not contain instructions to download or execute code from remote servers. Interactions with external services are abstracted through managed sub-agents (
moshu-researcher,moshu-architect) within the platform's execution environment. - [DATA_EXFILTRATION]: There are no indicators of attempts to access sensitive system paths (e.g.,
.ssh,.aws) or hardcoded credentials. The data handled by the skill is purely creative narrative content.
Audit Metadata