AI-assisted dataset annotation with COCO export — bbox, SAM2, DINOv3 methods
67%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Low
Low-risk findings worth noting
Fix and improve this skill with Tessl
tessl review fix ./skills/annotation/dataset-annotation/SKILL.mdThe skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
The skill runtime path reads outsider-authored free text from JSONL messages provided over stdin (`for line in sys.stdin` + `json.loads(line)`), and these messages can include free-form fields like `name`/`description`/`category` that the code embeds into the COCO output JSON (and would also be available to any LLM-driven components built around this pipeline).
SharpAI/DeepCamera
933dcc7
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.