[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7EV5J844X7BSRH6CSGKBTMD":3,"skill-same:01M7EV5J844X7BSRH6CSGKBTMD":47},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":25,"commitSha":26,"skillDirectory":27,"forksCount":28,"repositoryHtmlUrl":29,"repositoryHomepageUrl":6,"ownerHtmlUrl":30,"ownerBlogUrl":31,"ownerDescription":6,"updatedAt":32,"install":33,"resources":46},"01M7EV5J844X7BSRH6CSGKBTMD","ocr-pro",null,"Optical character recognition pipelines — engine choice, preprocessing, and accuracy — use when extracting text from images or scans.","TanNhatCMS\u002Fjarvis","TanNhatCMS","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F113796420?v=4","01M753MV17KG5CD3152G3WQEQY","KietNT","01M7EKEDYVYQ6PWVJJMZ911VMX",0,"MIT","https:\u002F\u002Fgithub.com\u002FTanNhatCMS\u002Fjarvis\u002Fblob\u002Fmain\u002Fskills\u002Fams-ocr-pro\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: ocr-pro\ndescription: Optical character recognition pipelines — engine choice, preprocessing, and accuracy — use when extracting text from images or scans.\ncategory: document-processing\n---\n\n## Overview\n\nOCR turns images of text — scans, photos, screenshots — into machine-readable\ntext. Accuracy depends less on the engine than on input quality and post-\nprocessing. This skill covers choosing an OCR approach, preprocessing images\nfor maximum accuracy, and validating results.\n\n## When to use\n\n- Digitizing scanned documents, PDFs-as-images, or paper archives\n- Extracting text from photos (receipts, signs, whiteboards, IDs)\n- Choosing between local OCR engines and cloud vision APIs\n- Improving accuracy on difficult inputs (handwriting, low contrast, skew)\n- Building validation for OCR output in a pipeline\n\n## Core concepts\n\n**Garbage in, garbage out — preprocessing is half the job.** Denoise, deskew,\nincrease contrast, and upscale small text (300 DPI equivalent is the rule of\nthumb) before OCR. A 30-second preprocessing pass often beats switching engines.\n\n**Engine choice is a trade-off.** Local engines (Tesseract and its successors)\nare free, private, and good for clean printed text. Cloud vision APIs handle\nhandwriting, complex layouts, and poor photos better but cost money and send\ndata off-site. Neural\u002Fdocument-AI services add layout understanding (tables,\nforms, key-value pairs) — worth it for invoices and forms.\n\n**Layout analysis precedes recognition.** Good OCR segments the page into\nblocks, lines, and words first. Multi-column pages, tables, and mixed\norientations need layout-aware processing — naive line-by-line OCR interleaves\ncolumns.\n\n**Confidence scores are signal.** Engines report per-word\u002Fper-character\nconfidence. Use it: route low-confidence segments to human review, and never\nlet sub-threshold numbers flow into financial or legal records unchecked.\n\n**Language and domain tuning.** Set the expected language(s); add custom\ndictionaries or patterns for domain vocabulary (product codes, medical terms).\nConstraining the character set (digits only for amounts) dramatically cuts\ncharacter-confusion errors (0\u002FO, 1\u002Fl\u002FI).\n\n## Practical workflow\n\n1. **Assess the input:** scan quality, language(s), layout complexity,\n   handwriting presence — this determines engine and preprocessing.\n2. **Build the preprocessing chain:** grayscale → denoise → deskew →\n   contrast\u002Fthreshold → upscale to ~300 DPI; inspect intermediate outputs on\n   samples.\n3. **Run OCR with the right config:** page segmentation mode matched to layout\n   (single block vs sparse text vs single line), language packs installed,\n   character whitelist where applicable.\n4. **Post-process:** spell\u002Fdomain correction, regex validation of structured\n   fields (dates, amounts, IDs), and format normalization.\n5. **Validate with ground truth:** hand-transcribe a sample set, measure\n   character\u002Fword error rates, and set an acceptance threshold before\n   production.\n6. **Design the human-in-the-loop path:** low-confidence items queue for\n   review; corrections feed back into dictionaries or fine-tuning data.\n\n## Common pitfalls\n\n- **Skipping preprocessing** and blaming the engine — most accuracy problems\n  are input problems.\n- **Trusting OCR numbers blindly** in financial\u002Flegal contexts — always\n  validate with checksums, totals, or human review above a risk threshold.\n- **Wrong segmentation mode** — the single most common Tesseract misconfiguration;\n  match it to the actual layout.\n- **Ignoring rotation\u002Fskew** — even 2–3 degrees of skew degrades accuracy\n  measurably; deskew first.\n- **Sending sensitive documents to cloud OCR** without considering data\n  residency and privacy obligations — PII in images is still PII.\n- **No handling for multi-page documents** — page ordering, headers\u002Ffooters\n  repeated per page, and tables spanning pages all need explicit logic.\n","## Overview\n\nOCR turns images of text — scans, photos, screenshots — into machine-readable\ntext. Accuracy depends less on the engine than on input quality and post-\nprocessing. This skill covers choosing an OCR approach, preprocessing images\nfor maximum accuracy, and validating results.\n\n## When to use\n\n- Digitizing scanned documents, PDFs-as-images, or paper archives\n- Extracting text from photos (receipts, signs, whiteboards, IDs)\n- Choosing between local OCR engines and cloud vision APIs\n- Improving accuracy on difficult inputs (handwriting, low contrast, skew)\n- Building validation for OCR output in a pipeline\n\n## Core concepts\n\n**Garbage in, garbage out — preprocessing is half the job.** Denoise, deskew,\nincrease contrast, and upscale small text (300 DPI equivalent is the rule of\nthumb) before OCR. A 30-second preprocessing pass often beats switching engines.\n\n**Engine choice is a trade-off.** Local engines (Tesseract and its successors)\nare free, private, and good for clean printed text. Cloud vision APIs handle\nhandwriting, complex layouts, and poor photos better but cost money and send\ndata off-site. Neural\u002Fdocument-AI services add layout understanding (tables,\nforms, key-value pairs) — worth it for invoices and forms.\n\n**Layout analysis precedes recognition.** Good OCR segments the page into\nblocks, lines, and words first. Multi-column pages, tables, and mixed\norientations need layout-aware processing — naive line-by-line OCR interleaves\ncolumns.\n\n**Confidence scores are signal.** Engines report per-word\u002Fper-character\nconfidence. Use it: route low-confidence segments to human review, and never\nlet sub-threshold numbers flow into financial or legal records unchecked.\n\n**Language and domain tuning.** Set the expected language(s); add custom\ndictionaries or patterns for domain vocabulary (product codes, medical terms).\nConstraining the character set (digits only for amounts) dramatically cuts\ncharacter-confusion errors (0\u002FO, 1\u002Fl\u002FI).\n\n## Practical workflow\n\n1. **Assess the input:** scan quality, language(s), layout complexity,\n   handwriting presence — this determines engine and preprocessing.\n2. **Build the preprocessing chain:** grayscale → denoise → deskew →\n   contrast\u002Fthreshold → upscale to ~300 DPI; inspect intermediate outputs on\n   samples.\n3. **Run OCR with the right config:** page segmentation mode matched to layout\n   (single block vs sparse text vs single line), language packs installed,\n   character whitelist where applicable.\n4. **Post-process:** spell\u002Fdomain correction, regex validation of structured\n   fields (dates, amounts, IDs), and format normalization.\n5. **Validate with ground truth:** hand-transcribe a sample set, measure\n   character\u002Fword error rates, and set an acceptance threshold before\n   production.\n6. **Design the human-in-the-loop path:** low-confidence items queue for\n   review; corrections feed back into dictionaries or fine-tuning data.\n\n## Common pitfalls\n\n- **Skipping preprocessing** and blaming the engine — most accuracy problems\n  are input problems.\n- **Trusting OCR numbers blindly** in financial\u002Flegal contexts — always\n  validate with checksums, totals, or human review above a risk threshold.\n- **Wrong segmentation mode** — the single most common Tesseract misconfiguration;\n  match it to the actual layout.\n- **Ignoring rotation\u002Fskew** — even 2–3 degrees of skew degrades accuracy\n  measurably; deskew first.\n- **Sending sensitive documents to cloud OCR** without considering data\n  residency and privacy obligations — PII in images is still PII.\n- **No handling for multi-page documents** — page ordering, headers\u002Ffooters\n  repeated per page, and tables spanning pages all need explicit logic.\n",{"name":5,"category":24,"description":7},"document-processing","main","af388a77c2fe5e0b4fd879cca4300a0ffa9a6579","skills\u002Fams-ocr-pro",1,"https:\u002F\u002Fgithub.com\u002FTanNhatCMS\u002Fjarvis","https:\u002F\u002Fgithub.com\u002FTanNhatCMS","https:\u002F\u002Fmrkiet.name.vn","2026-11-07 14:23:08",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":45},[35,38,41],{"tool":36,"command":37},"npx","npx skills add TanNhatCMS\u002Fjarvis --skill ocr-pro",{"tool":39,"command":40},"bunx","bunx skills add TanNhatCMS\u002Fjarvis --skill ocr-pro",{"tool":42,"command":43},"pnpm","pnpm dlx skills add TanNhatCMS\u002Fjarvis --skill ocr-pro","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FTanNhatCMS\u002Fjarvis\u002Fblob\u002Fmain\u002Fskills\u002Fams-ocr-pro\u002FSKILL.md",true,[],[48,55,63,70,77,84],{"skillNo":49,"name":50,"nameZh":6,"description":51,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":52,"categories":53,"keywords":54,"publishedAt":19},"01M7EWYJR0J8DARY7E59Z6Y9MV","web-scrape","Turn web pages into LLM-ready markdown via the Firecrawl API (renders JS, anti-bot handling). Use when browser_open\u002Ffetch fails on JS-heavy pages, or when you need clean markdown\u002Fstructured extraction from a site. Requires FIRECRAWL_API_KEY.","https:\u002F\u002Fgithub.com\u002FTanNhatCMS\u002Fjarvis\u002Fblob\u002Fmain\u002Fskills\u002Fweb-scrape\u002FSKILL.md",[],[],{"skillNo":56,"name":57,"nameZh":6,"description":58,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":59,"sourceHtmlUrl":60,"categories":61,"keywords":62,"publishedAt":19},"01M7EWYBN5P6F2AJG3FR24T9BP","design-taste-frontend","tasteskill v2, the anti-slop frontend skill. Art direction for websites, distilled from a study of 247 curated gallery sites (One Page Love, Awwwards Sites of the Day). Deals researched directions (idea, composition, type, palette, photographs, section blueprints) from a library, enforces a strict no-slop house style with a real-browser inspector, and runs a look-and-fix loop on screenshots. Use this whenever you build, design, redesign, restyle, polish or review a landing page, marketing site, portfolio, product page, one-pager, hero or any public-facing web page, in any stack (HTML\u002FCSS, React, Next.js, Vue, Svelte, Astro, Tailwind) — also when the user says a page looks generic, templated, \"AI-made\", sloppy, or asks to make it look designed, premium, award-level or less like every other site. Not for dashboards or app screens.","https:\u002F\u002Fgithub.com\u002FTanNhatCMS\u002Fjarvis\u002Fblob\u002Fmain\u002Fskills\u002Ftaste-design\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FTanNhatCMS\u002Fjarvis\u002Fblob\u002Fmain\u002Fskills\u002Ftaste-design\u002FSKILL.md",[],[],{"skillNo":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":67,"categories":68,"keywords":69,"publishedAt":19},"01M7EWY8FQFNQ2GB7NZZWGSSDX","test-driven-development","Use when implementing any feature or bugfix, before writing implementation code","https:\u002F\u002Fgithub.com\u002FTanNhatCMS\u002Fjarvis\u002Fblob\u002Fmain\u002Fskills\u002Fsuperpowers-tdd\u002FSKILL.md",[],[],{"skillNo":71,"name":72,"nameZh":6,"description":73,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":19},"01M7EWY5SHBFRH1HDHVWM9K7TV","systematic-debugging","Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes","https:\u002F\u002Fgithub.com\u002FTanNhatCMS\u002Fjarvis\u002Fblob\u002Fmain\u002Fskills\u002Fsuperpowers-debugging\u002FSKILL.md",[],[],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":19},"01M7EWY3NM0Z0JAKS7HDZ24K66","requesting-code-review","Use when completing tasks, implementing major features, or before merging to verify work meets requirements","https:\u002F\u002Fgithub.com\u002FTanNhatCMS\u002Fjarvis\u002Fblob\u002Fmain\u002Fskills\u002Fsuperpowers-code-review\u002FSKILL.md",[],[],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":19},"01M7EWY0WRMWX4M9ZFY0WKC593","brainstorming","You MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation).","https:\u002F\u002Fgithub.com\u002FTanNhatCMS\u002Fjarvis\u002Fblob\u002Fmain\u002Fskills\u002Fsuperpowers-brainstorming\u002FSKILL.md",[],[]]