Skill Registry

Find the right skill for the job your agent is doing.

Search by task, platform, trust profile, or GitHub adoption. OpenAgentSkill keeps the interface human-readable and the API ready for agents.

2
Shown
42
Tracks
92
Targets

Decision filters

Choose by scenario, quality, and trust signals.

Showing 2 skills matching "instance-segmentation"

Best blend of quality, stars, freshness, and agent usage

1

Ultralytics

VERIFIEDEXCELLENT · 100TRUST · 93SAFE · REVIEWEDDESIGN

Ultralytics YOLO 🚀

$ npx skills add ultralytics/ultralytics
58.5K stars77 quality93 trustReviewed with permission notesClaude CodePushed todaySafe to try
QualityHigh-confidence pick with strong adoption and healthy maintenance signals.
TrustStrong OpenAgentSkill Trust Score across adoption, recent maintenance, license clarity, documentation, dependency risk, and install availability.Review: Documentation summary is thin
Safety gateUsable candidate, but the agent should surface permission and audit notes before installation.Action: Require human approval before installing into a real workspace.

Scenario

Multimodal media

I need my agent to process images, video, or audio and extract useful information.

Agents

Claude Code + CLI + Codex

4 install targets

Maintenance

fresh

Pushed today

Risk

Safe to try

Documentation summary is thin

Gate

review

Reviewed with permission notes

pythonmachine-learning
by ultralyticsQuick view
2

Supervision

VERIFIEDEXCELLENT · 100TRUST · 95SAFE · VERIFIEDDESIGN

We write your reusable computer vision tools. 💜

$ npx skills add roboflow/supervision
44.4K stars76 quality95 trustVerifiedClaude CodePushed todaySafe to try
QualityHigh-confidence pick with strong adoption and healthy maintenance signals.
TrustStrong OpenAgentSkill Trust Score across adoption, recent maintenance, license clarity, documentation, dependency risk, and install availability.
Safety gateStrong metadata, audit, install, and review signals. Suitable for agent shortlists after normal workspace review.Action: Allow agent install in a sandbox or low-risk workspace, then promote after one successful narrow task.

Scenario

Multimodal media

I need my agent to process images, video, or audio and extract useful information.

Agents

Claude Code + CLI + Codex

4 install targets

Maintenance

fresh

Pushed today

Risk

Safe to try

No major risk signals from available metadata

Gate

allow

Verified

pythonmachine-learning
by roboflowQuick view