CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

pdf

Use this skill whenever the user wants to do anything with PDF files. This includes reading or extracting text/tables from PDFs, combining or merging multiple PDFs into one, splitting PDFs apart, rotating pages, adding watermarks, creating new PDFs, filling PDF forms, encrypting/decrypting PDFs, extracting images, and OCR on scanned PDFs to make them searchable. If the user mentions a .pdf file or asks to produce one, use this skill.

Quality

88%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

High

Do not use without reviewing

SKILL.md
Quality
Evals
Security

Quality

Content

72%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.

A highly actionable, code-dense reference whose examples are copy-paste ready, but it has two structural weaknesses: batch operations (merge/split/encrypt) lack any validation or verification steps, and progressive disclosure is broken in practice — the referenced REFERENCE.md and FORMS.md do not exist in the bundle while the scripts/ directory is never mentioned. Trimming duplicate tool coverage would also tighten token efficiency.

Suggestions

Add validation checkpoints after batch operations — e.g., after splitting, verify the output file count matches len(reader.pages); after encrypting, reopen the output with the password to confirm it is readable.

Fix the reference layer: either include REFERENCE.md and FORMS.md in the bundle or remove the pointers, and add a section in SKILL.md listing the scripts/ helpers (e.g., fill_fillable_fields.py, extract_form_structure.py) so they are discoverable.

Consolidate redundant tool coverage — pick one canonical merge/split/rotate approach (pypdf or qpdf) and move the alternates (pdftk 'if available', advanced table extraction to Excel) into the reference file to reduce token cost.

DimensionReasoningScore

Conciseness

The body is almost pure executable code with headers and near-zero explanation of concepts Claude already knows — no 'what a PDF is' padding — so it sits above the 'minor instances of over-explanation' anchor. It falls short of 5 due to redundancy: three merge approaches (pypdf, qpdf, pdftk 'if available'), a split example that repeats the merge loop pattern, and a Quick Reference table that restates sections.

4 / 5

Actionability

Every section is complete, copy-paste-ready, executable code or commands covering the common cases (merge, split, extract text/tables, create, rotate, watermark, encrypt, OCR, image extraction), with dependencies noted where needed ('# Requires: pip install pytesseract pdf2image'). This matches the 'fully executable; copy-paste ready; covers the common cases' anchor exactly.

5 / 5

Workflow Clarity

Each recipe is a single unambiguous action, but the skill includes batch operations — merging multiple PDFs, splitting into one file per page, encrypting — and none carry validation or verification steps (e.g., checking output page counts or that the encrypted file opens). The rubric's guideline explicitly caps workflow clarity at 3 for batch operations without validation, so it cannot score 4 despite the clear per-task clarity.

3 / 5

Progressive Disclosure

Structure is present and references are clearly signaled (Overview and Next Steps point to REFERENCE.md and FORMS.md), but scored against the actual bundle: both referenced files are missing from the skill directory, so the pointers are broken, and the 8 scripts in scripts/ (e.g., fill_fillable_fields.py, extract_form_structure.py) are never surfaced from SKILL.md — they are undiscoverable. Additionally ~300 lines of recipes are inlined where the skill's own REFERENCE.md design implies a split, matching the 'content that should be separate is inline' anchor.

3 / 5

Total

15

/

20

Passed

Description

100%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.

This is an exemplary description: concrete and comprehensive capability listing, natural trigger vocabulary including the .pdf extension, explicit what-and-when guidance, and a clearly distinct PDF niche. It closely mirrors the rubric's own good examples.

DimensionReasoningScore

Specificity

The description enumerates many concrete, comprehensive actions — 'reading or extracting text/tables', 'combining or merging', 'splitting PDFs apart', 'rotating pages', 'adding watermarks', 'filling PDF forms', 'encrypting/decrypting', 'extracting images', and 'OCR on scanned PDFs' — matching the anchor for multiple specific concrete actions with comprehensive coverage; nothing is vague or generic.

5 / 5

Completeness

It explicitly answers both what (the enumerated action list) and when ('Use this skill whenever the user wants to do anything with PDF files... If the user mentions a .pdf file or asks to produce one, use this skill'), with concrete trigger phrases — a direct match for the top anchor. It is not a 4 because the 'when' is fully explicit rather than needing more specificity.

5 / 5

Trigger Term Quality

It covers natural terms users would say plus synonyms and the file extension: 'PDF files', 'PDFs', '.pdf file', 'merging', 'splitting', 'filling PDF forms', 'OCR'. This matches the comprehensive anchor ('PDF files, PDFs, forms, document extraction, .pdf') — no common variation is missing.

5 / 5

Distinctiveness Conflict Risk

The description is firmly anchored to the PDF niche with distinct triggers ('mentions a .pdf file', 'asks to produce one'), giving minimal overlap risk with non-PDF skills — matching the 'clear niche with distinct triggers' anchor. The phrase 'anything with PDF files' is broad in scope but stays within a distinct domain, so it does not fall to 4.

5 / 5

Total

20

/

20

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

UnicomAI/wanwu

Path

configs/microservice/bff-service/configs/agent-skills/anthropics/pdf/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.