Skill audit report
OS Agent Survey audit report.
This is the repo for the paper "OS Agents: A Survey on MLLM-based Agents for Computer, Phone and Browser Use" (ACL 2025 Oral).
OpenAgentSkill Trust Score
Stars, maintenance, license, docs, install safety, permission surface, and installability.
The Trust Score is OpenAgentSkill's adoption layer. It is designed to help an agent decide whether a skill is safe enough to shortlist before installation.
GitHub adoption
INFO62
485 GitHub stars
Stars/forks activity
WARN57
485 stars, 21 forks; issue activity unavailable in current metadata
Recent maintenance
INFO62
10mo since push
License clarity
WARN42
Unknown
README/SKILL.md completeness
PASS90
Metadata includes enough usage and workflow context
Dependency/runtime risk
PASS82
network or browser surface
Install availability
PASS92
npx skills add OS-Agent-Survey/OS-Agent-Survey
Install command safety
PASS92
standard package or runtime install path
Permission surface
INFO72
filesystem or document access, network or browser access
Repository evidence
PASS86
https://github.com/OS-Agent-Survey/OS-Agent-Survey
Review status
PASS88
AI review data available
Checks
Install and adoption review
Install path
92
npx skills add OS-Agent-Survey/OS-Agent-Survey
Repository
88
https://github.com/OS-Agent-Survey/OS-Agent-Survey
License
45
Unknown
Maintenance
62
10mo since push
AI review
88
Approved with no listed issues
README/SKILL.md completeness
90
Usable description available
Dependency risk
82
network or browser surface
Install command safety
92
standard package or runtime install path
Permission surface
72
filesystem or document access, network or browser access
Stars/forks activity
57
485 stars, 21 forks; issue activity unavailable in current metadata
Adoption
68
485 GitHub stars
Warnings
- License is unclear
- Quality score needs review
- Stars/forks activity: 485 stars, 21 forks; issue activity unavailable in current metadata
- License clarity: Unknown
Method
This report combines public metadata, AI review output, repository freshness, install readiness, OpenAgentSkill events, quality scoring, trust checks, and the agent safety gate. It is not a full source-code security review.
Compare nearby options