roboflow-data-management
Use when uploading images, labeling, organizing datasets, creating Roboflow projects (detection/segmentation/keypoint/classification), tags, splits, versions, or RoboQL search.
77
66%
Does it follow best practices?
Impact
94%
1.40xAverage score across 3 eval scenarios
Advisory
Suggest reviewing before use
Optimize this skill with Tessl
npx tessl skill review --optimize ./skills/data-management/SKILL.mdQuality
Discovery
82%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
The description excels at trigger term coverage and distinctiveness, clearly identifying Roboflow-specific workflows and terminology. However, it lacks an explicit 'what this skill does' statement—it jumps straight into 'Use when...' without first describing the skill's purpose (e.g., 'Manages computer vision datasets and projects on Roboflow'). Adding a brief capability summary before the trigger clause would improve completeness.
Suggestions
Add a leading capability statement before the 'Use when' clause, e.g., 'Manages computer vision datasets and projects on the Roboflow platform. Use when...'
Consider briefly mentioning the output or value provided (e.g., 'automates dataset preparation for training computer vision models') to strengthen the 'what does this do' aspect
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: uploading images, labeling, organizing datasets, creating Roboflow projects with specific types (detection/segmentation/keypoint/classification), tags, splits, versions, and RoboQL search. | 3 / 3 |
Completeness | The description starts with 'Use when' which addresses the 'when' aspect, but the 'what does this do' part is only implied through the trigger conditions rather than explicitly stated. It reads more like a list of triggers than a description of capabilities followed by explicit trigger guidance. | 2 / 3 |
Trigger Term Quality | Includes strong natural keywords users would say: 'uploading images', 'labeling', 'datasets', 'Roboflow', 'detection', 'segmentation', 'classification', 'keypoint', 'tags', 'splits', 'versions', and the platform-specific 'RoboQL search'. Good coverage of terms a user working with Roboflow would naturally use. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive due to the Roboflow-specific terminology (Roboflow projects, RoboQL search) and the specific project types (detection/segmentation/keypoint/classification). Very unlikely to conflict with other skills. | 3 / 3 |
Total | 11 / 12 Passed |
Implementation
50%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a comprehensive reference document for Roboflow data management that is well-organized with tables and clear sections. Its main weaknesses are that it reads more like documentation than an actionable skill — it describes features rather than providing step-by-step executable workflows with validation checkpoints. The content could benefit from being split into a concise overview skill with references to detailed sub-files for preprocessing, augmentation, and RoboQL reference tables.
Suggestions
Add concrete MCP tool invocation examples with parameters (e.g., a complete create_project -> upload -> generate_version workflow) to improve actionability.
Add validation/verification steps for destructive operations: check class list before deletion, verify version generation output, confirm upload counts match expectations.
Split the preprocessing options table, augmentation options table, and RoboQL filter reference into separate reference files, keeping only the most common options inline to improve progressive disclosure and conciseness.
Remove or condense explanatory text that Claude can infer (e.g., 'Tags are free-form labels on images for organization and filtering') and focus on the non-obvious constraints and gotchas.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is mostly efficient with good use of tables, but includes some content Claude already knows or could infer (e.g., explaining what tags are, what splits are, general concepts like 'frozen snapshot'). The augmentation and preprocessing tables are comprehensive but borderline verbose for a skill file — some of this reference material could be offloaded. The opening meta-note about plugin vs MCP resources is useful but lengthy. | 2 / 3 |
Actionability | The skill provides concrete CLI commands and a Python SDK snippet, plus specific MCP tool names. However, most guidance is descriptive/reference rather than executable — the MCP tool table lists tools but doesn't show example invocations with parameters. The RoboQL section gives good filter examples but lacks a complete workflow example showing how to chain tools together for a real task. | 2 / 3 |
Workflow Clarity | The version creation pipeline is clearly sequenced (5 steps), which is good. However, there are no validation checkpoints or error recovery steps for destructive operations like class changes (which are noted as irreversible) or version generation. The upload workflow lacks verification steps. The class management section warns about irreversibility but doesn't provide a safe workflow pattern. | 2 / 3 |
Progressive Disclosure | There is one cross-reference to the labeling skill at the bottom, and the content is organized with clear headers and tables. However, the file is quite long (~200+ lines of dense reference material) with preprocessing options, augmentation tables, and analytics details that could be split into separate reference files. For a skill file, this is borderline monolithic despite good internal structure. | 2 / 3 |
Total | 8 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- Repository
- roboflow/computer-vision-skills
- Commit
02936d5
- Reviewed
Table of Contents
Is this your skill?
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.