smart-ocr

Extract text from images and scanned documents using PaddleOCR - supports 100+ languages

Quality

55%

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

Fix and improve this skill with Tessl

tessl review fix ./.trae/openclaw-skills/smart-ocr/SKILL.md

Quality

Content

57%Weight 40%Scale 1-5

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

The body is highly actionable with broad executable code coverage, but it is verbose and monolithic: API reference material and redundant examples are inlined rather than split into reference files, and batch/destructive workflows lack validation checkpoints.

Suggestions

Add validation to batch and PDF workflows: verify OCR results (e.g. confidence threshold, non-empty text) before cleanup, and handle per-page/per-image errors so one failure doesn't abort the batch.

Move the supported-languages dictionary, the full configuration-options reference, and the three worked examples into separate reference files (e.g. LANGUAGES.md, CONFIG.md, EXAMPLES.md) linked from the overview to improve progressive disclosure.

Tighten redundancy: the three Examples re-demonstrate PaddleOCR initialization and result iteration already shown in Fundamentals; consolidate or reference the canonical pattern instead of repeating it.

Dimension	Reasoning	Score
Conciseness	The body is mostly useful executable code, but the inlined supported-languages dictionary, a full configuration-options dump of every parameter, and three worked examples that re-demonstrate the same init/iterate pattern could be trimmed, matching anchor 3.	3 / 5
Actionability	Concrete, copy-paste-ready Python covers images, PDFs, URLs/bytes, batch, and multilingual cases; minor gaps such as the PDF example calling os.remove without importing os and batch_ocr relying on a global ocr keep it just below anchor 5.	4 / 5
Workflow Clarity	Steps are present ('How to Use') but batch operations (batch_ocr, PDF temp-file processing with os.remove) lack validation or verification checkpoints, so workflow clarity is capped at 3 per the batch/destructive-operations rule.	3 / 5
Progressive Disclosure	Section headers give reasonable navigation, but ~460 lines are fully inlined with configuration reference and language-list content that belongs in separate files, and there are no external file references at all, placing it at anchor 3.	3 / 5
	Total	13 / 20 Passed

Description

53%Weight 40%Scale 1-5

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 clearly states what the skill does and names a distinctive library, but it omits any explicit 'when to use' trigger guidance and is missing several natural trigger terms (OCR, screenshots, PDF). It is a competent but incomplete description.

Suggestions

Add an explicit trigger clause, e.g. 'Use when the user wants to OCR an image, read text from a photo or screenshot, or extract text from a scanned PDF.'

Broaden trigger terms to include 'OCR', 'screenshots', 'PDF', 'photos', and common file extensions (.png, .jpg, .pdf).

List one or two more concrete actions (e.g. 'extract text with position and confidence data') to lift specificity toward anchor 4.

Dimension	Reasoning	Score
Specificity	Names the OCR domain and one concrete action ('Extract text from images and scanned documents') plus source types, but does not list several specific actions, so it sits at anchor 3 rather than 4.	3 / 5
Completeness	Has a clear 'what' ('Extract text from images and scanned documents using PaddleOCR') but no 'Use when...' trigger clause, so completeness is capped at 3 per the missing-trigger-guidance rule.	3 / 5
Trigger Term Quality	Includes relevant natural terms ('images', 'scanned documents', 'text', 'languages') but omits common variations users say like 'OCR', 'screenshots', 'PDF', 'photos', and file extensions, matching anchor 3.	3 / 5
Distinctiveness Conflict Risk	The OCR-from-images niche is clear and PaddleOCR is a named, distinctive library with low conflict risk, but it lacks the explicit trigger phrases that would place it at anchor 5.	4 / 5
	Total	13 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: huangruiteng/CS-Notes
Path: .trae/openclaw-skills/smart-ocr/SKILL.md
Commit: 43eb317

Reviewed: 3 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.