Supervision
Roboflow's reusable computer vision utilities for annotation, tracking, and visualising detection model outputs.
Supervision
Supervision by Roboflow is a library of reusable computer-vision utilities that sits on top of any detection or segmentation model. It provides annotators, trackers, zone counting, and dataset tools that eliminate boilerplate when building vision pipelines.
Key Features
- Model-agnostic
Detectionsdataclass compatible with YOLO, SAM, Detectron2, and more - Rich annotation suite: BoundingBox, Mask, Label, Halo, Ellipse, Corner, Trace annotators
- Object tracking with ByteTrack, SORT, and BotSort with one-line integration
- Zone counting, line crossing counters, and polygon zones
- Dataset tools: load COCO/YOLO/Pascal VOC, split, merge, and export
Quick Start
pip install supervision
import supervision as sv
from ultralytics import YOLO # or any detector
model = YOLO("yolov8n.pt")
image = sv.load_image("image.jpg")
results = model(image)[0]
detections = sv.Detections.from_ultralytics(results)
annotator = sv.BoxAnnotator()
annotated = annotator.annotate(scene=image.copy(), detections=detections)
sv.plot_image(annotated)
npx ai-supply add supervision-vision-toolkit
Curated mirror of the open-source Supervision (MIT). Get it from the source.
Compromise signals — malicious or tampered code (leaked secrets, backdoors, a dropped executable) — reduce the score, and known dependency CVEs carry a bounded penalty (they warrant review but never QUARANTINE — update the dependency to clear). Other dangerous-by-capability traits are risk surface, expected for some capabilities. Every finding is mapped to its OWASP control below.
Findings mapped to the OWASP Top 10 for LLM Applications (2025) and the OWASP Machine Learning Security Top 10. Expand any flagged control for the exact findings — compromise reduces the score; expected/risk-surface do not, except a known CVE, which carries a small bounded penalty (high/critical → Review).
The same gate an agent runs before installing (POST /api/v1/trust/supervision-vision-toolkit/check). Click a policy:
Consume Supervision programmatically. Authenticate with an API key or session — see Authorize an agent.
# Agents: CHECK BEFORE YOU INSTALL (no auth) — score, grade, level, capability manifest
curl https://ai-supply.store/api/v1/trust/supervision-vision-toolkit
# Gate against your org policy (returns { pass, violations })
curl -X POST https://ai-supply.store/api/v1/trust/supervision-vision-toolkit/check \
-H "Content-Type: application/json" \
-d '{"minGrade":"B","denyPermissions":["shell"],"denyUnknownEgress":true}'
# CLI
npx ai-supply add supervision-vision-toolkit
# REST (install → download)
curl -X POST https://ai-supply.store/api/v1/listings/supervision-vision-toolkit/install \
-H "Authorization: Bearer $AIM_KEY"
# MCP tool
install_listing({ "slug": "supervision-vision-toolkit" })OpenAPI spec →Curated mirror — latest upstream source. See the repository for tagged releases.