PDF

Before you start: read this whole install prompt, tell me in a few lines what this skill does and what installing it involves, and wait for my go-ahead before you do anything else. Prerequisites — Source-available licence warning first: this skill is NOT Apache-2.0 and not open source. Its own LICENSE.txt (read it in the folder) says use of these materials is governed by my agreement with Anthropic for its services (or, with no separate agreement, Anthropic's Consumer or Commercial Terms of Service), and its additional restrictions include not retaining copies of these materials outside the Services, not distributing them to any third party, and not creating derivative works. Installing this folder copies the materials onto my machine: only proceed if my own Anthropic agreement covers that, and stop and tell me if that is not clear. No account, key or payment is needed to copy the files, and no companion script is run during installation — the Python scripts in the folder are copied as-is only. Never paste any PDF password into this install: encrypting or decrypting my own PDFs is a use-time decision I make later, file by file. At use time (not during install) the skill needs Python with pypdf, pdfplumber and reportlab, the Poppler tools and qpdf, all running locally on my own files; OCR, only if I ask for it, additionally needs pytesseract and pdf2image. Install "PDF" for me. It covers anything done with a PDF file: reading and extracting text and tables with pypdf and pdfplumber, merging, splitting, rotating and watermarking, creating PDFs with reportlab, encrypting and decrypting, extracting images with Poppler tools, OCR on scanned PDFs by converting pages to images first, and filling PDF forms by following its forms guide step by step — checking whether fields are fillable, extracting the field structure, rendering validation images, then writing the values in. Its folder holds the main guide, two longer companion guides (forms and an advanced reference), a proprietary LICENSE.txt, and eight small Python scripts that support the form-filling workflow. Repository: https://github.com/anthropics/skills (the skill is the whole skills/pdf/ folder at the pinned commit: SKILL.md plus 11 companion files, 12 files in total) Pinned version: install the code at commit dbd4588f9e1033efb41dad4bef2f7947c8993d44 exactly. Do not substitute a newer version. If this commit is unreachable, stop and tell me. 1. Clone https://github.com/anthropics/skills into a temporary folder, check out commit dbd4588f9e1033efb41dad4bef2f7947c8993d44, and copy the skills/pdf/ folder from that checkout. Then compare every file in the folder with the fingerprints recorded for this listing on 2026-10-22: [FINGERPRINTS] - LICENSE.txt: SHA-256 79f6d8f5b427252fa3b1c11ecdbdb6bf610b944f7530b4de78f770f38741cfaa - SKILL.md: SHA-256 9f78b8359fbd4943ad260a7a1e436e5a96503406d6c34e99f69223d647d85b9c - forms.md: SHA-256 9530b3f57034792e0242a3abfe8b3c296466329551916f2ff6c5a70db78f7e44 - reference.md: SHA-256 03a5f964f8abecbbe156f363356e927e864d7ee964f1012c84ee1bfc8acbeb95 - scripts/check_bounding_boxes.py: SHA-256 0ced522402dd7b3c82a0be9518cf4e485fa9cc0467d7beab6ddb04cb36fcc3e1 - scripts/check_fillable_fields.py: SHA-256 1fe10b9980a5493c9bbec8c20630c603eaa475b45403751c08f26738736647b8 - scripts/convert_pdf_to_images.py: SHA-256 7f58e292a67e18a8ff000d30bad2469389546731f7e8a698cc1d0aed001ea15d - scripts/create_validation_image.py: SHA-256 8fa9fd7962c9f940045b81cf6b991e7c1057c1838e3f9cc6d305007a222ffebf - scripts/extract_form_field_info.py: SHA-256 bb235b36f497bbb1c7c6821cecf527bbba7213520de11219da99f7b13a38c109 - scripts/extract_form_structure.py: SHA-256 6814e3fe8f78c2b16d766c132b5a727ff7a31b71d76a7d46bd6e40fe50bc97c5 - scripts/fill_fillable_fields.py: SHA-256 d140872c2e92e41b1e05370ae8d270c5b7f5ebdf55d9905f03bc974a9c708ff5 - scripts/fill_pdf_form_with_annotations.py: SHA-256 8a56e063a49ed53c0b0e1d312ecf8023af76816ed139486ec3b539873de202fd If a file does not match, it changed since the listing was prepared: do not install it yet, tell me what you found and let me decide. 2. Read SKILL.md in full, and tell me in plain words: what the skill does when used, what it connects to, whether it reads or sends any personal information or credentials, and anything that seems unusual. Stop on any red flag and tell me before going further. Confirm these points from the files: - The folder holds 12 files: SKILL.md, LICENSE.txt (proprietary, Anthropic terms), forms.md and reference.md (companion guides), and 8 small Python scripts under scripts/ that detect, extract and fill PDF form fields and convert pages to validation images. Nothing in the folder is run during installation; the scripts are copied as-is. - Nothing is sent anywhere during installation itself; the only download at install time is the skill folder from GitHub at the pinned commit. No PDF of mine is opened, read or sent during installation, and no PDF password is asked for, read or stored. - At use time everything runs locally on my own PDF files with the Python libraries and command-line tools named in the prerequisites. The instructions name no third-party service to send documents to; the one operation that touches file protection, encrypting or decrypting a PDF, uses a password I supply at that moment for my own file. - The licence point from the prerequisites is part of the safety picture: these materials are source-available under my Anthropic agreement with the restrictions in LICENSE.txt, not Apache-2.0. Do not present them as open source, and do not redistribute the installed folder. - Do not run any companion script, do not open any PDF of mine, and do not use the skill in any way during this install: using it is a separate step that only starts when I ask for it later. 3. When I say "yes, install", install it as a skill: copy the whole pdf folder into the agent's skills directory, in a folder named exactly "pdf", keeping its 12 files in their original relative layout (SKILL.md at its root). 4. Verify with no network calls: frontmatter valid (name: pdf), the 12 files in place with matching fingerprints. Do not run anything from the folder and do not use the skill to verify. 5. Report what was installed, where, and what I still need to do myself (for example: that using it needs local Python libraries (pypdf, pdfplumber, reportlab) plus Poppler and qpdf, that its files are source-available under my Anthropic agreement (not Apache-2.0) with the restrictions in its LICENSE.txt, and that its companion scripts run only at use time, on my own files, locally). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.