Adding .Claude folder - #2295
nrshrivatsan wants to merge 3 commits into
Conversation
|
vtsn seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account. You have signed the CLA already but the status is still pending? Let us recheck it. |
Borda
left a comment
Borda
left a comment
There was a problem hiding this comment.
@nrshrivatsan, it seems like an interesting contribution. Could you share your motivation or why AGENTS?MD and extended contribution guide is not sufficient?
|
Idea is simple - Supervision could be of immense value to non coders such as
However the barrier to entry could be lowered if they could use Claude code to jump start ( instead of cold start) Hence the idea. Hope this makes sense. |
|
This is great to hear just to undestand they would most likely use supervison as a tool not develop features for supervison, right? |
|
Yes- tool based use is the target access pattern. |
|
Any updates on this PR? |
Still figuring out the benefit compare to the CALUDE. md |
|
BTW, how was this generated? |
There was a problem hiding this comment.
Pull request overview
Adds Claude Code guidance for setup, exploration, recipes, and production pipelines.
Changes:
- Adds four proposed skills with examples.
- Catalogs local cookbook notebooks.
- Provides supervision workflow guidance.
Assessment: Code quality 1/5 · Testing 1/5 · Documentation 2/5
Reviewed changes
Copilot reviewed 4 out of 4 changed files in this pull request and generated 9 comments.
| File | Description |
|---|---|
.claude/skills/use.md |
Video, tracking, analytics, and SAHI guidance. |
.claude/skills/setup.md |
Environment setup guidance. |
.claude/skills/explore.md |
Detection and dataset exploration guidance. |
.claude/skills/cookbooks.md |
Local notebook catalog. |
Suppressed comments (1)
.claude/skills/use.md:83
- This zone example cannot run: the escaped brackets are invalid Python, and
PolygonZone.__init__does not acceptframe_resolution_wh, causing aTypeErrorafter the syntax is corrected. The current API derives its mask resolution from the polygon.
polygon = np.array(\[[10, 10], [100, 10], [100, 100], [10, 100]\])
zone = sv.PolygonZone(polygon=polygon, frame_resolution_wh=video_info.resolution_wh)
💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.
| @@ -0,0 +1,120 @@ | |||
| <skill> | |||
| @@ -0,0 +1,54 @@ | |||
| <skill> | |||
| @@ -0,0 +1,123 @@ | |||
| <skill> | |||
| @@ -0,0 +1,44 @@ | |||
| <skill> | |||
| for frame in frame_generator: | ||
| \# 1. Model Inference | ||
| \# 2. sv.Detections | ||
| \# 3. Annotate | ||
| sink.write_frame(annotated_frame) |
| import supervision as sv | ||
|
|
||
| def callback(image: np.ndarray) -> sv.Detections: | ||
| \# Your model inference logic here | ||
| \# Must return sv.Detections object | ||
| return sv.Detections(...) |
| try: | ||
| import inference | ||
| print("✅ inference installed") | ||
| except ImportError: | ||
| print("❌ inference missing") |
| labels = \[ | ||
| f"{class_id} {confidence:.2f}" | ||
| for class_id, confidence | ||
| in zip(detections.class_id, detections.confidence) | ||
| \] |
| for tracker_id in detections.tracker_id: | ||
| if is_in_zone\[detections.tracker_id == tracker_id\]: | ||
| on_entry(tracker_id) |
What
Claude code with skills for users to get started