drawio-reconstruction

Reconstructs reference images into high-fidelity, editable Draw.io files with rendered previews: native Draw.io elements carry text and structure, SVG covers simple icons that match the reference, and cropped or transparent PNGs preserve complex visuals. Use when the user wants a diagram image, research figure, architecture diagram, slide, UI screenshot, or image folder turned into `.drawio` XML; batch requests use a manifest with bounded parallelism when available and a full-fidelity serial fallback otherwise.

Quality

88%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

77%Weight 40%Scale 1-3

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The content is highly actionable with strong, validated workflows and real bundled scripts, but it is a dense single-file monolith with redundant restatements of crop/icon guidance that could be tightened or split into reference files. Strengths: executable commands and explicit audit feedback loops.

Suggestions

Consolidate the overlapping crop/icon guidance now spread across 'Reconstruction Strategy', 'Medium Decision Rules', 'Screenshot Crop Rules', and 'Icon Reconstruction Discipline' into one section to cut redundancy and tokens.

Move the detailed medium/crop/seam rules into a references/ file (e.g., MEDIUM_RULES.md) and keep SKILL.md as a concise overview that links one level deep, improving progressive disclosure.

Add the missing requirements.txt bundle file that crop_assist.py instructions reference, or remove that install instruction, so every referenced path resolves.

Dimension	Reasoning	Score
Conciseness	The body largely assumes Claude's competence and adds genuinely novel guidance (medium decision rules, crop/seam discipline), but at ~350 lines it restates similar crop/icon rules across Reconstruction Strategy, Medium Decision Rules, Screenshot Crop Rules, and Icon Reconstruction Discipline — material that could be tightened. Not the bottom anchor (no basic-concept padding), but above-average verbosity keeps it off the lean 3.	2 / 3
Actionability	Provides concrete, copy-paste-ready commands with real flags ('$SKILL_DIR/scripts/crop_assist.py ... --roi x,y,w,h --anchor ... --exclude ...') and verification commands; all referenced scripts exist in ./scripts/. Fully executable guidance matching the top anchor.	3 / 3
Workflow Clarity	Sequenced Default Workflow (11 steps) and Batch Workflow with explicit validation checkpoints — 'Treat any layout failure as a blocking defect', run check_drawio.py/batch_verify.py, then a full visual audit and iterate until both technical checks and audit pass — giving the feedback loops the rubric rewards for batch/destructive operations.	3 / 3
Progressive Disclosure	A monolithic ~350-line SKILL.md with all detail inline and no separate reference files (no references/ directory); sections are well-organized by headers and the bundled scripts are correctly referenced one-level-deep via $SKILL_DIR, but content that should be split (detailed medium/crop rules) is inline rather than in reference docs.	2 / 3
	Total	10 / 12 Passed

Description

100%Weight 40%Scale 1-3

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 is exemplary: third-person voice, concrete capabilities, broad natural trigger terms, and an explicit 'Use when' clause covering both single-image and batch cases. It avoids fluff and over-claims.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions and outputs: 'native Draw.io elements carry text and structure, SVG covers simple icons ... and cropped or transparent PNGs preserve complex visuals', plus manifest-driven batch handling — matching the rubric's top anchor for multiple specific concrete actions.	3 / 3
Completeness	Explicitly answers both what (reconstructs images into editable .drawio with previews) and when via a clear 'Use when the user wants ...' trigger clause, hitting the top anchor.	3 / 3
Trigger Term Quality	Covers natural user terms a person would actually say — 'diagram image, research figure, architecture diagram, slide, UI screenshot, or image folder' and '.drawio XML' — giving good coverage rather than jargon.	3 / 3
Distinctiveness Conflict Risk	Occupies a clear niche (image→drawio reconstruction) with distinct triggers and output format, making it unlikely to fire for unrelated skills.	3 / 3
	Total	12 / 12 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: HKUSTDial/Supervisor-Skills
Path: skills/drawio-reconstruction/SKILL.md
Commit: aff5de9

Reviewed: 23 days ago

Table of Contents

Discovery Implementation Validation

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.