What does this skill do, and when should you use it?
This is a universal Agent Skill for the Nutrient DWS API, working with Claude Code, Codex, Gemini CLI, Cursor, and 40+ other agents. Through pinned Python helper scripts it calls the cloud-hosted Nutrient DWS Processor to generate and convert PDFs, assemble packets, run OCR, extract text/tables/key-value pairs, apply watermarks and staged redactions, sign, password-protect, optimize, and produce PDF/A or PDF/UA output. It includes a product router (separating Processor, Data Extraction, and Accessibility), a paid-run approval gate, and a two-stage redaction flow, with strong emphasis on key safety and explicit confirmation for irreversible operations. It suits users who need high-fidelity document processing and accept uploading files to a paid cloud API.
- Reads local files and calls the Nutrient DWS Processor API, producing conversions to PDF, Office formats, images, HTML, or Markdown
- Performs page-level operations: merge, split, add/delete/duplicate pages, rotate
- Runs multi-language OCR on scans and extracts text, tables, and key-value pairs
- Provides pattern-based and AI two-stage redaction: stage annotations for visual review, then apply irreversibly after separate approval
- Handles CMS/CAdES digital signing, form filling (Instant JSON/XFDF), text watermarks, optimization, and linearization
- Targets PDF/A and PDF/UA compliance output with validation guidance
- A user with a scanned contract who needs OCR first, then full-text extraction to a file
- HR or admin staff filling an employee onboarding PDF form and digitally signing it for archival
- Legal/compliance teams redacting SSNs, emails, and credit card numbers before sharing patient records
- Developers converting report.docx to PDF or generating official PDFs from HTML templates in a pipeline
- Archivists converting PDFs to PDF/A-2a for long-term storage and validating compliance per the skill's guidance
- Publishers optimizing and linearizing large PDFs for fast web delivery
- Workflows where document content must not leave your machine — every invocation sends files to the Nutrient DWS cloud API and consumes credits
- Casual individual users who cannot obtain a Nutrient API key or prefer not to pay per use
- Locked-down environments without Python 3.10+ and uv installed
How do you install this skill?
- Core function depends entirely on the overseas paid api.nutrient.io service; mainland-China reachability is unverified, and a paid API key with credit consumption is required
- Remote-URL inputs make Nutrient's server fetch the URL — guard against internal-network addresses; manually verify transferred files and credit estimates before every paid run
- All safety rules (approval gate, two-stage redaction, key hygiene) rely on the agent following instructions with no technical sandbox; use only in a trusted agent environment
- Publisher is unverified; this is a static source review with no executed API calls — real conversion/signing/redaction output quality is unverified
- Shell / CLI
- Network access
- Local filesystem
Python 3.10+uvnutrient-dws==3.1.0Nutrient DWS API key (NUTRIENT_API_KEY)
Requires Python 3.10+, uv, and a Nutrient API key. Use npx skills to install to all detected agents (recommended), target specific agents, or install globally; or manually copy the skill folder into your agent's skills directory.
All agents (npx skills, recommended)
npx skills add PSPDFKit-labs/nutrient-agent-skillSpecific agents only
npx skills add PSPDFKit-labs/nutrient-agent-skill -a claude-code -a codex -a cursorGlobal install
npx skills add PSPDFKit-labs/nutrient-agent-skill -gClaude Code (manual)
git clone https://github.com/PSPDFKit-labs/nutrient-agent-skill.git
cp -r nutrient-agent-skill/nutrient-document-processing ~/.claude/skills/How do you use this skill?
Once installed, send your agent any of these to trigger it:
- Convert report.docx to PDF
- OCR this German scan and extract the text to a file
- Redact all social security numbers, emails, and credit card numbers from patient-records.pdf
- Add a DRAFT watermark to proposal.pdf
Once NUTRIENT_API_KEY is configured, simply ask your agent in natural language to process a document. The skill routes to the right product, picks a pinned scripts/ helper, or copies the multi-step workflow from custom-workflow-template.py. Before each paid run the agent must present the operation, files transferred, and a credit estimate, then wait for your approval before passing --estimated-credits and --confirm-external-processing. Outputs are written atomically, refuse to overwrite existing paths, and retries require a new path.
What are this skill's strengths and limitations?
- Broad coverage: conversion, OCR, redaction, signing, forms, and compliance output in one skill, compatible with 40+ agents
- Thoughtful security design: API keys verified for presence only, passwords read solely from chmod 600 local files, atomic non-overwriting output, mandatory approval before paid runs
- Two-stage redaction with visual review and verification; signed PDFs treated as immutable; compliance outputs come with validation requirements
- Pinned dependencies (nutrient-dws==3.1.0 via PEP 723) prevent contract drift
- Every invocation sends document content to an external API and consumes credits — cost grows with usage and requires per-run approval
- Heavy prerequisites: Nutrient API key, Python 3.10+, and uv
- Repo-level license is unknown in GitHub metadata (SKILL.md and LICENSE.txt are MIT-0)
- The Data Extraction and Accessibility products are not implemented inside the skill; users must consult their current official clients/docs separately
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Nutrient Document Processing Skill this page | 69 · Recommended | ★ 18 | 1mo ago | — |
| Markdown to EPUB Skill | 52 · Use with care | ★ 167 | 1y ago | MIT |
| NeMo Retriever Document Retrieval Skill ✓ NVIDIA · Official | 41 · Not recommended | ★ 3.6k | 11d ago | Apache-2.0 |
| Open Cowork PDF Skill | 69 · Recommended | ★ 2.2k | 12d ago | MIT |
| Image to Editable PPT | 54 · Use with care | ★ 2.9k | 1mo ago | MIT |
Compared with @nutrient-sdk/dws-mcp-server (the npm MCP server shown in the README badges), this skill uses the Agent Skills standard rather than the MCP protocol; the README lists both, positioning this skill as an alternative integration path to the same Nutrient DWS capabilities.
How did FollowSkills review this skill?
Evidence shows excellent least-privilege and confirmation design: every script embeds a paid-run approval gate (--confirm-external-processing with credit estimate), the key lives only in the environment and is never printed, passwords are read only from chmod-600 files, AI redaction is two-stage with separately approved irreversible apply, outputs are atomic and refuse overwrite, and CI writes tokens to 600-permission temp files. Deducted because external-transfer enforcement relies on the agent following instructions rather than hard technical enforcement, remote-URL inputs carry SSRF-style risk mitigated only by guidance, and there is no sandbox isolation.
Offline contract tests cover page-range parsing, all three typed output shapes, overwrite refusal, secret-file permissions, per-invocation confirmation, --help without a key, and pre-network blocking when confirmation is missing; CI workflows execute them, meeting the static uplift condition. Deducted because real API call paths (DWS 3.1.0 client behavior, failure feedback quality) cannot be statically verified and depend on nutrient-dws==3.1.0 availability and correctness.
The product router cleanly separates Processor / Data Extraction / Accessibility; trigger description and non-fit ranges are explicit; setup and runtime requirements are complete. Deducted because core function depends entirely on the overseas paid Nutrient DWS API (api.nutrient.io) with questionable mainland-China reachability, requires a paid API key, and offers no Chinese-language support notes.
Layered architecture (SKILL.md router plus modular references), MIT-0 license, version 2.0.0, release workflow with changelog and pinned source commit, human publication gates. Deducted because publisher is unverified, maintenance responsibility and update path are only implicit in repo structure, and the testing guide is a manual procedure rather than automated user-facing validation.
The skill covers core document-processing scenarios; the pinned typed client reduces hand-written-payload errors, and safe output writers plus the approval gate should yield directly usable artifacts. Deducted because static review cannot verify real output correctness/completeness, value depends on a paid API (cost/benefit scales with usage), and README marketing demos are not independently checkable.
Real CI workflows plus committed test suites covering key paths (subprocess-level blocking tests, payload fixture comparison, absence of unsafe raw payloads) exceed claim-only evidence. Deducted because this is a static read without execution, tests cover only offline contracts and source structure, and real API behavior plus third-party corroboration are missing.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →