CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

ocr-and-documents

Extract and structure text from scanned PDFs and document images with a local-first OCR workflow, layout-aware fallbacks, and explicit heavyweight dependency controls.

Quality

80%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./packages/ekko-agent/skills/ocr-and-documents/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

93%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 an exemplary lean skill: a lightest-first decision ladder, executable commands verified against real bundled scripts, careful heavyweight-dependency gating, and an explicit verification checklist. The only notable gap is the absence of an error-recovery feedback loop tying verification failures back to corrective re-extraction.

DimensionReasoningScore

Conciseness

The body never explains concepts Claude already knows (no "what is OCR" or library primers), delegates exact CLI options to "Use --help for exact options", and confines marker-pdf's cost warning to two sentences of necessary policy. Every section is instruction or constraint, matching "lean and efficient; every token earns its place".

5 / 5

Actionability

The helper invocations ("python3 <baseDirectory>/scripts/extract_pymupdf.py input.pdf -o extracted.md", "extract_marker.py --check", "python3 -m pip install pymupdf") are copy-paste ready, and the <baseDirectory> placeholder is explicitly resolved by the "Start here" instruction to call skill_view. Both referenced scripts exist in the bundle and their flags (--check, --markdown, -o) match the documented behavior.

5 / 5

Workflow Clarity

The "Choose the lightest path" ladder gives a clear lightest-to-heaviest sequence with explicit conditions per rung, and the Verification section is an explicit checklist (compare against rendered pages, verify table alignment, mark uncertain characters). It stops short of anchor 5 because there is no explicit validate-then-fix feedback loop, e.g. what to do when verification reveals misaligned tables or garbled text.

4 / 5

Progressive Disclosure

The body is ~44 lines with well-organized sections, and both referenced paths (scripts/extract_pymupdf.py, scripts/extract_marker.py) exist one level deep in the bundle; CLI bulk is correctly externalized into those scripts rather than inlined. Under the under-50-lines guideline, this earns the top score for well-organized structure with no external-reference needs beyond the scripts.

5 / 5

Total

19

/

20

Passed

Description

61%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 communicates a clear, fairly distinctive niche (OCR for scanned/image-based documents) with a couple of concrete actions and natural trigger terms. Its main weaknesses are the missing explicit "Use when..." trigger clause, which caps completeness, and trailing approach descriptors that add specificity of tone but not of action.

Suggestions

Add an explicit trigger clause, e.g. "Use when a PDF or document is scanned or image-based, or when the user mentions OCR or extracting text from an image", to raise completeness.

Replace buzzword-style qualifiers ("layout-aware fallbacks", "explicit heavyweight dependency controls") with concrete actions the bundled scripts already support, such as extracting tables and embedded images.

Include common natural variations such as "image text extraction" or ".pdf" to broaden trigger-term coverage.

DimensionReasoningScore

Specificity

"Extract and structure text from scanned PDFs and document images" names the domain plus two concrete actions, but the trailing phrases "layout-aware fallbacks" and "explicit heavyweight dependency controls" describe an approach rather than adding actions, and capabilities the scripts actually have (tables, images, metadata) are absent. This matches 'names domain and 1-2 concrete actions, but not comprehensive'; a 4 would require several specific listed actions.

3 / 5

Completeness

The "what" is clear ("Extract and structure text from scanned PDFs and document images") but there is no "Use when..." clause or equivalent explicit trigger guidance, which per the judging guidelines caps completeness at 3. The "when" is only weakly implied by the mention of scanned sources.

3 / 5

Trigger Term Quality

"scanned PDFs", "document images", "OCR", and "extract" are phrases users naturally say when they need this skill. Coverage is good but misses common variations such as "image text extraction" and file extensions like ".pdf", so it falls short of the comprehensive-synonym anchor at 5.

4 / 5

Distinctiveness Conflict Risk

"scanned PDFs and document images" combined with an OCR workflow carves a mostly distinct niche with minor overlap risk against generic PDF-extraction skills (a scanned PDF is still a PDF file). It is far more specific than the anchor-3 example ("Works with document files") but lacks the fully distinct triggers of anchor 5.

4 / 5

Total

14

/

20

Passed

Validation

87%

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

Validation — 14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

metadata_field

'metadata' should map string keys to string values

Warning

Total

14

/

16

Passed

Repository

EKKOLearnAI/hermes-studio

Path

packages/ekko-agent/skills/ocr-and-documents/SKILL.md

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.