{ "generated": "2026-08-18T18:09:46.801Z", "source": "entries.yaml", "api": { "base": "https://toolshed.lemon-agent.dev", "check": "https://toolshed.lemon-agent.dev/check?from=&to=", "convert": "https://toolshed.lemon-agent.dev/convert/", "free_tier_daily": 10, "note": "Entries with a hosted block run on our server. Every one is free to try — 10 conversions per caller (per IP) per UTC day, reported in an x-free-tier-remaining header — and priced per call past that: HTTP 402 with an x402 envelope, or HTTP 429 with a Retry-After while payment is switched off. Entries without a hosted block are local-only references." }, "entries": [ { "id": "markdown-to-pdf", "x": "Markdown (with math, code blocks, citations)", "y": "PDF", "xlabel": "markdown", "ylabel": "pdf", "tool": "pandoc", "kind": "deterministic", "verdict": "Pandoc plus a Typst or LaTeX engine gives the same PDF on every run, with real pagination, footnotes and cross-references. The naive alternative — print-to-PDF from a Markdown previewer — loses footnote links and repaginates differently on every machine, which shows up as a diff you can't reproduce.", "hosted": null, "local": { "tool": "pandoc", "install": "brew install pandoc typst" }, "install": "brew install pandoc typst", "url": "https://pandoc.org", "caveats": "PDF output needs an external engine; Typst is fast and light, LaTeX (texlive) is heavy but still the only option for some journal templates.", "escalate": "Never for the conversion. A model is only useful upstream — writing or restructuring the Markdown before pandoc touches it.", "verified": "2026-08-18" }, { "id": "markdown-to-docx", "x": "Markdown", "y": "DOCX styled to a house template", "xlabel": "markdown", "ylabel": "docx", "tool": "pandoc", "kind": "deterministic", "verdict": "`--reference-doc=template.docx` maps headings, tables and code blocks onto a real Word style sheet, so a human reviewer gets an editable document instead of a wall of Normal-styled text. Pasting rendered Markdown into Word is the failure mode: styles arrive as direct formatting and the first edit shatters them.", "hosted": null, "local": { "tool": "pandoc", "install": "brew install pandoc" }, "install": "brew install pandoc", "url": "https://pandoc.org", "caveats": "Word-native features (comments, tracked changes, floats with anchors) don't round-trip; you get clean structure, not pixel-matched design.", "escalate": "Never — style mapping is a template problem, not a judgment problem. Rewriting content for a different audience is a separate upstream step.", "verified": "2026-08-18" }, { "id": "docx-to-markdown", "x": "DOCX", "y": "Markdown + extracted media", "xlabel": "docx", "ylabel": "markdown", "tool": "pandoc", "kind": "deterministic", "verdict": "Pandoc reads the DOCX XML directly, so headings, lists, tables and footnotes land as real Markdown structure, and `--extract-media` writes embedded images to disk instead of dropping them. Pasting a DOCX into a model to \"convert it\" is the failure mode: you pay tokens for a plausible-looking file with quietly renumbered lists and vanished tables.", "hosted": null, "local": { "tool": "pandoc", "install": "brew install pandoc" }, "install": "brew install pandoc", "url": "https://pandoc.org", "caveats": "Tracked changes and comments are discarded unless you pass `--track-changes=all`; pseudo-headings faked with bold text stay paragraphs, because they were never structure.", "escalate": "Only when there is no structure to recover — a document where every heading was faked with formatting and you want semantic headings inferred.", "verified": "2026-08-18" }, { "id": "html-markdown", "x": "Saved HTML page / static HTML file", "y": "Markdown", "xlabel": "html", "ylabel": "markdown", "tool": "pandoc", "kind": "deterministic", "verdict": "For HTML already on disk, pandoc's reader is exact and fast, and `--wrap=none` keeps the output diff-friendly. Feeding a whole page to a model to \"clean it up\" burns tokens and invents heading levels; pandoc preserves precisely the structure that is there. The hosted endpoint runs Turndown instead of pandoc — same job, no install — free to try inside the daily free tier, then a priced x402 call like every other hosted tool here.", "hosted": { "path": "/convert/html-markdown", "price": { "amount_usd": 0.001, "scheme": "exact" }, "status": "live", "free_tier_daily": 10 }, "local": { "tool": "pandoc", "install": "brew install pandoc" }, "install": "brew install pandoc", "url": "https://pandoc.org", "caveats": "Pandoc converts everything, including nav, ads and cookie banners — pre-strip with a readability-style extractor (trafilatura, readability-lxml) if you only want the article body. The hosted endpoint has the same blind spot.", "escalate": "Never for the markup. Boilerplate removal is a heuristic problem first; a model is a last resort for pages where the article body is genuinely ambiguous.", "verified": "2026-08-18" }, { "id": "md-html", "x": "Markdown", "y": "HTML", "xlabel": "markdown", "ylabel": "html", "tool": "pandoc", "kind": "deterministic", "verdict": "Markdown to HTML is a solved parse, so the only real questions are which dialect and which extensions. Pandoc is the local answer and handles tables, footnotes and math; the hosted endpoint runs `marked` over CommonMark plus GitHub tables, which is what most agents actually want when they say \"render this Markdown\".", "hosted": { "path": "/convert/md-html", "price": { "amount_usd": 0.001, "scheme": "exact" }, "status": "live", "free_tier_daily": 10 }, "local": { "tool": "pandoc", "install": "brew install pandoc" }, "install": "brew install pandoc", "url": "https://pandoc.org", "caveats": "The hosted endpoint emits an HTML fragment, not a full document — no ``, no stylesheet. It does not sanitize: raw HTML in the input passes through, so do not render untrusted output into a page without sanitizing it yourself.", "escalate": "Never.", "verified": "2026-08-18" }, { "id": "rendered-dom-to-markdown", "x": "Live DOM in a browser or headless page (JS-rendered)", "y": "Markdown", "xlabel": "rendered dom", "ylabel": "markdown", "tool": "turndown", "kind": "deterministic", "verdict": "When the content only exists after JavaScript runs, the conversion belongs where the DOM is — Turndown runs in-page or in Node against a parsed DOM and takes custom rules for app-specific markup. Fetching the raw HTML and running a static converter gets you an empty shell.", "hosted": null, "local": { "tool": "turndown", "install": "npm install turndown" }, "install": "npm install turndown", "url": "https://github.com/mixmark-io/turndown", "caveats": "Smaller rule set than pandoc; tables need the GFM plugin and you own the escaping edge cases.", "escalate": "Never for markup. Use a model only to decide which region of a hostile app UI is \"the content\" — then hand that node to turndown.", "verified": "2026-08-18" }, { "id": "markdown-to-epub", "x": "Markdown (multi-chapter manuscript)", "y": "EPUB 3", "xlabel": "markdown", "ylabel": "epub", "tool": "pandoc", "kind": "deterministic", "verdict": "`pandoc -o book.epub` produces a valid EPUB with a real table of contents, metadata and chapter splitting via `--split-level`; readers care about that structure far more than typography. The naive path — export to PDF and sideload — gives a fixed-layout file that is unreadable on a six-inch screen.", "hosted": null, "local": { "tool": "pandoc", "install": "brew install pandoc" }, "install": "brew install pandoc", "url": "https://pandoc.org", "caveats": "Run epubcheck before publishing; store-specific requirements (Apple, Kobo) go beyond mere validity.", "escalate": "Never.", "verified": "2026-08-18" }, { "id": "video-to-portable-mp4", "x": "Arbitrary video (any container/codec)", "y": "H.264 MP4 that plays everywhere", "xlabel": "video", "ylabel": "mp4", "tool": "ffmpeg", "kind": "deterministic", "verdict": "`-c:v libx264 -crf 20 -preset slow -c:a aac -movflags +faststart` is the boring correct answer for delivery: CRF targets quality rather than a guessed bitrate, and faststart lets playback begin before the file finishes downloading. GUI \"convert to MP4\" presets routinely omit faststart, so web playback stalls until the whole file lands.", "hosted": null, "local": { "tool": "ffmpeg", "install": "brew install ffmpeg" }, "install": "brew install ffmpeg", "url": "https://ffmpeg.org", "caveats": "CRF values are codec-relative — don't reuse an x264 number for x265 or AV1. Hardware encoders are much faster and meaningfully worse per bit.", "escalate": "Never. Encoding is math; the only judgment is your quality/size target, which is a knob you set once.", "verified": "2026-08-18" }, { "id": "video-to-audio-track", "x": "Video file with an audio track", "y": "Audio file (m4a/wav)", "xlabel": "video", "ylabel": "audio file", "tool": "ffmpeg", "kind": "deterministic", "verdict": "`ffmpeg -i in.mp4 -vn -c:a copy out.m4a` lifts the existing audio stream with no re-encode and no generation loss; switch to `-c:a pcm_s16le out.wav` only when the next tool demands PCM. Reflexively re-encoding to MP3 adds a lossy generation for nothing.", "hosted": null, "local": { "tool": "ffmpeg", "install": "brew install ffmpeg" }, "install": "brew install ffmpeg", "url": "https://ffmpeg.org", "caveats": "Stream copy needs a container that accepts the codec (AAC to .m4a, not .wav); multi-track sources need an explicit `-map`.", "escalate": "Never.", "verified": "2026-08-18" }, { "id": "audio-transcode-and-normalize", "x": "WAV / FLAC / arbitrary audio", "y": "MP3 or Opus at a fixed sample rate and loudness", "xlabel": "audio", "ylabel": "mp3 / opus", "tool": "ffmpeg", "kind": "deterministic", "verdict": "ffmpeg does the format change, the resample (`-ar`) and loudness normalization (`loudnorm`) in a single pass, which matters because chaining separate tools resamples twice and compounds artifacts. For speech and web delivery Opus at a low bitrate is clearly better than MP3 at the same size.", "hosted": null, "local": { "tool": "ffmpeg", "install": "brew install ffmpeg" }, "install": "brew install ffmpeg", "url": "https://ffmpeg.org", "caveats": "Single-pass loudnorm is approximate; use the two-pass measure-then-apply form when you actually care about hitting a LUFS target.", "escalate": "Never.", "verified": "2026-08-18" }, { "id": "heic-to-jpeg", "x": "HEIC/HEIF photos from an iPhone", "y": "JPEG", "xlabel": "heic photos", "ylabel": "jpeg", "tool": "imagemagick", "kind": "deterministic", "verdict": "`magick mogrify -format jpg -quality 88 *.heic` converts a folder in one line and keeps EXIF, which matters when the timestamps are the reason you kept the photos. Re-exporting through a photo app or a web converter tends to strip or rewrite metadata — and uploading personal photos to a random site is a privacy trade you probably didn't intend to make.", "hosted": null, "local": { "tool": "imagemagick", "install": "brew install imagemagick libheif" }, "install": "brew install imagemagick libheif", "url": "https://imagemagick.org", "caveats": "Requires a libheif-enabled build (check `magick -list format | grep -i heic`); slower and hungrier than libvips on large batches.", "escalate": "Never.", "verified": "2026-08-18" }, { "id": "bulk-image-resize", "x": "Large batch of source images", "y": "Web-sized JPEG/WebP/AVIF derivatives", "xlabel": "images", "ylabel": "web images", "tool": "libvips", "kind": "deterministic", "verdict": "`vipsthumbnail` streams through a batch on a small memory footprint and gets the defaults right — shrink-on-load for JPEG, sane colour handling. The naive `magick -resize` loop over thousands of files is where the out-of-memory kills and the hour-long builds come from.", "hosted": null, "local": { "tool": "libvips", "install": "brew install vips" }, "install": "brew install vips", "url": "https://www.libvips.org", "caveats": "Fewer effects and filters than ImageMagick; heavy compositing or annotation work still belongs in ImageMagick, at its cost.", "escalate": "Never for the resize. Picking crop focal points across a large library is the one adjacent task where a model (saliency detection) earns its keep.", "verified": "2026-08-18" }, { "id": "svg-to-png", "x": "Static SVG", "y": "PNG at a chosen scale", "xlabel": "svg", "ylabel": "png", "tool": "resvg", "kind": "deterministic", "verdict": "resvg renders an SVG to PNG with no browser and no JS runtime, so it runs in CI and produces the same bytes run after run. Headless Chrome does the same job while pulling in a browser-sized dependency and a screenshot timing race.", "hosted": null, "local": { "tool": "resvg", "install": "cargo install resvg" }, "install": "cargo install resvg", "url": "https://github.com/linebender/resvg", "caveats": "Static SVG only — no scripting, no remote resource fetching, and web fonts must be installed locally. `rsvg-convert` (librsvg) is the distro-packaged alternative.", "escalate": "Never.", "verified": "2026-08-18" }, { "id": "json-yaml", "x": "JSON", "y": "YAML", "xlabel": "json", "ylabel": "yaml", "tool": "yq", "kind": "deterministic", "verdict": "mikefarah's yq speaks both formats natively (`yq -P`) and preserves comments through YAML edits, which is the whole reason to use it on real config files. The hosted endpoint is the no-install version for the common case: valid JSON in, block-style YAML out.", "hosted": { "path": "/convert/json-yaml", "price": { "amount_usd": 0.001, "scheme": "exact" }, "status": "live", "free_tier_daily": 10 }, "local": { "tool": "yq", "install": "brew install yq" }, "install": "brew install yq", "url": "https://github.com/mikefarah/yq", "caveats": "Two unrelated tools are named yq — the Go one (mikefarah) and a Python jq wrapper (kislyuk); the flags differ, so pin which one your scripts assume. The hosted endpoint has no comments to preserve, because JSON has none.", "escalate": "Never.", "verified": "2026-08-18" }, { "id": "yaml-json", "x": "YAML", "y": "JSON", "xlabel": "yaml", "ylabel": "json", "tool": "yq", "kind": "deterministic", "verdict": "`yq -o=json` is the local answer, and the direction that bites is this one: YAML is the larger language, so anchors, multi-document streams and unquoted `no`/`yes` all have to land somewhere in JSON. The hosted endpoint takes the first document of a stream and resolves anchors, which is what a config-reading agent wants.", "hosted": { "path": "/convert/yaml-json", "price": { "amount_usd": 0.001, "scheme": "exact" }, "status": "live", "free_tier_daily": 10 }, "local": { "tool": "yq", "install": "brew install yq" }, "install": "brew install yq", "url": "https://github.com/mikefarah/yq", "caveats": "Multi-document YAML collapses to its first document on the hosted endpoint — use local yq if you need all of them. YAML dates and other non-JSON scalar types are stringified.", "escalate": "Never.", "verified": "2026-08-18" }, { "id": "csv-json", "x": "CSV", "y": "JSON", "xlabel": "csv", "ylabel": "json", "tool": "csvkit", "kind": "deterministic", "verdict": "`csvjson data.csv` turns a header row plus records into an array of objects, which is the shape almost everything downstream wants. The hosted endpoint does the same thing with an RFC-4180 parser, so quoted commas and embedded newlines survive — which is exactly what a split-on-comma one-liner gets wrong.", "hosted": { "path": "/convert/csv-json", "price": { "amount_usd": 0.001, "scheme": "exact" }, "status": "live", "free_tier_daily": 10 }, "local": { "tool": "csvkit", "install": "pip install csvkit" }, "install": "pip install csvkit", "url": "https://csvkit.readthedocs.io", "caveats": "Every value comes out a string; the hosted endpoint does not guess types, because guessing is where leading zeros and long IDs get destroyed. Rows shorter than the header are padded with empty strings, and a row with more fields than the header is an error rather than a silent truncation.", "escalate": "Never for the parse. Deciding what the columns *mean* — see messy-csv-to-clean-csv — is the judgment half.", "verified": "2026-08-18" }, { "id": "json-reshape", "x": "Nested JSON from an API", "y": "Flat JSON / NDJSON / CSV rows", "xlabel": "json", "ylabel": "json / ndjson", "tool": "jq", "kind": "deterministic", "verdict": "jq is a real language for this: `-r`, `@csv`, `to_entries`, `group_by` cover most reshapes in a line, and the result is an auditable program you can re-run on tomorrow's payload. Asking a model to \"reshape this JSON\" produces an answer, not a program — it doesn't survive the next record and quietly guesses on nulls.", "hosted": null, "local": { "tool": "jq", "install": "brew install jq" }, "install": "brew install jq", "url": "https://jqlang.github.io/jq/", "caveats": "Very large documents need `--stream` or NDJSON input; the syntax has a genuine learning curve that people repeatedly underestimate.", "escalate": "Use a model to *write the jq program* when the shape is unfamiliar — then keep the program, drop the model, and put the program in version control.", "verified": "2026-08-18" }, { "id": "messy-csv-to-clean-csv", "x": "Messy CSV (ragged rows, BOM, mixed quoting, duplicate headers)", "y": "Clean, validated UTF-8 CSV", "xlabel": "messy csv", "ylabel": "clean csv", "tool": "qsv", "kind": "deterministic", "verdict": "`qsv input` normalizes quoting and line endings, and `validate`/`headers`/`dedup` tell you what is actually wrong before the data reaches anything downstream — all at streaming speed on files far past spreadsheet size. Opening it in Excel to \"fix it\" is the classic failure: leading zeros vanish, long IDs become floats, and dates get silently localized.", "hosted": null, "local": { "tool": "qsv", "install": "brew install qsv" }, "install": "brew install qsv", "url": "https://github.com/dathere/qsv", "caveats": "Some subcommands depend on how the binary was built (feature flags); csvkit is slower but pip-installable everywhere and reads more input formats.", "escalate": "Only when the fix requires meaning — deciding that \"N/A\", \"-\" and \"\" are the same null, or that two differently-spelled columns are one field. Extract deterministically, let a model propose the mapping, then encode that mapping as a rule you can re-run.", "verified": "2026-08-18" }, { "id": "xlsx-to-csv", "x": "XLSX workbook (specific sheet)", "y": "CSV", "xlabel": "xlsx", "ylabel": "csv", "tool": "csvkit", "kind": "deterministic", "verdict": "`in2csv --sheet \"Sheet2\" book.xlsx` gets the sheet you asked for with stable typing and scripts across a directory. Save-As-CSV in Excel exports only the active sheet, applies the machine's locale to dates and separators, and cannot be automated — which is how two people produce two different CSVs from one workbook.", "hosted": null, "local": { "tool": "csvkit", "install": "pip install csvkit" }, "install": "pip install csvkit", "url": "https://csvkit.readthedocs.io", "caveats": "Formulas export as their last cached value; merged cells and multi-row headers still need reshaping. Very large workbooks are slow — `qsv excel` is the fast path.", "escalate": "When the sheet is a human-formatted report rather than a table (title rows, merged headers, stacked sub-tables), identifying the header row is genuine judgment: dump deterministically first, then let a model name the header row and the sub-table boundaries.", "verified": "2026-08-18" }, { "id": "csv-to-sqlite", "x": "Large CSV", "y": "Queryable SQLite database", "xlabel": "csv", "ylabel": "sqlite", "tool": "sqlite3", "kind": "deterministic", "verdict": "`.import --csv data.csv t` buys you indexes, joins and SQL over a file too big for a spreadsheet, with no service and no ORM. Loading a multi-hundred-megabyte CSV into pandas just to filter it is the naive alternative — paying RAM for work the disk should be doing.", "hosted": null, "local": { "tool": "sqlite3", "install": "built-in (macOS/most Linux); brew install sqlite for a current version" }, "install": "built-in (macOS/most Linux); brew install sqlite for a current version", "url": "https://sqlite.org/cli.html", "caveats": "`.import` types every column as TEXT unless you create the table first — declare the schema when numeric comparison matters. csvkit's `csvsql --db` infers types at a large speed cost.", "escalate": "Never.", "verified": "2026-08-18" }, { "id": "sqlite-to-json", "x": "SQLite query result", "y": "JSON / NDJSON for an API or an agent", "xlabel": "sqlite", "ylabel": "json / ndjson", "tool": "sqlite3", "kind": "deterministic", "verdict": "`sqlite3 -json db.sqlite \"select ...\"` (or `.mode json`) emits valid JSON straight from the CLI, composes with jq in a pipe, and needs no serialization layer. Hand-rolling a Python export script for this creates code you now have to maintain and test.", "hosted": null, "local": { "tool": "sqlite3", "install": "built-in" }, "install": "built-in", "url": "https://sqlite.org/cli.html", "caveats": "Needs a reasonably modern SQLite (3.33+) for the `-json` flag; check BLOB and NULL handling before shipping the output anywhere.", "escalate": "Never.", "verified": "2026-08-18" }, { "id": "pdf-digital-born-to-text", "x": "PDF (digital-born — has a text layer)", "y": "Plain text / layout-preserved text", "xlabel": "digital-born pdf", "ylabel": "plain text", "tool": "pdftotext", "kind": "deterministic", "verdict": "If `pdftotext -layout` returns real text, the job is already finished — the characters are in the file and extraction is exact and instant. This is the entry that governs the whole PDF branch of the directory: run pdftotext first on every PDF, and only when it comes back empty or garbled does the file deserve OCR or a model.", "hosted": null, "local": { "tool": "pdftotext", "install": "brew install poppler" }, "install": "brew install poppler", "url": "https://poppler.freedesktop.org", "caveats": "Multi-column and tabular layouts still interleave; `-layout` helps, and `-bbox-layout` gives coordinates when you need to reconstruct columns yourself.", "escalate": "Never for extraction. Escalate only for interpretation — turning extracted text into a schema (see messy-doc-to-schema).", "verified": "2026-08-18" }, { "id": "pdf-scanned-to-searchable", "x": "PDF (scanned — page images, no text layer)", "y": "Searchable PDF + extractable text", "xlabel": "scanned pdf", "ylabel": "searchable pdf", "tool": "ocrmypdf", "kind": "hybrid", "verdict": "ocrmypdf adds an invisible text layer to the original pages, so the file stays a faithful scan and becomes greppable and citable; `--skip-text` and `--redo-ocr` make it safe to sweep a mixed directory. Pointing raw tesseract at the page images throws the PDF away and leaves you a text blob nobody can cite a page number from.", "hosted": null, "local": { "tool": "ocrmypdf", "install": "brew install ocrmypdf" }, "install": "brew install ocrmypdf", "url": "https://ocrmypdf.readthedocs.io", "caveats": "OCR quality is bounded by scan quality — deskew and clean help, but phone photos of curved pages stay bad. Slow and CPU-hungry on long documents.", "escalate": "When accuracy matters on hard scans — handwriting, historical type, dense tables — a vision model reads them better. Still run OCR first and keep its output as a cross-check: OCR fails loudly, models fail by writing plausible words.", "verified": "2026-08-18" }, { "id": "pdf-to-page-images", "x": "PDF page(s)", "y": "PNG/JPEG page images", "xlabel": "pdf", "ylabel": "page images", "tool": "pdftoppm", "kind": "deterministic", "verdict": "`pdftoppm -r 200 -png` renders pages reproducibly and is the correct pre-step whenever OCR or a vision model is the next stage, because you control DPI, page range and colour — which is exactly what downstream accuracy depends on. Screenshotting pages by hand introduces cropping and scaling nobody wrote down.", "hosted": null, "local": { "tool": "pdftoppm", "install": "brew install poppler" }, "install": "brew install poppler", "url": "https://poppler.freedesktop.org", "caveats": "300+ dpi over a long document produces very large files quickly; pick the lowest DPI the downstream stage tolerates and record it.", "escalate": "This pair is the deterministic half of a model workflow — the model comes after, on the rendered images, never instead of the render.", "verified": "2026-08-18" }, { "id": "pdf-tables-to-csv", "x": "PDF with ruled or whitespace-aligned tables", "y": "CSV / DataFrame", "xlabel": "pdf tables", "ylabel": "csv", "tool": "camelot", "kind": "deterministic", "verdict": "Camelot's lattice mode reconstructs cells from the table's actual ruling lines, so on a bordered table it is exact rather than approximate; stream mode handles whitespace-aligned tables with more supervision. Copy-pasting a table out of a PDF viewer merges columns unpredictably and you won't notice until a number is wrong in a report.", "hosted": null, "local": { "tool": "camelot", "install": "pip install \"camelot-py[base]\"" }, "install": "pip install \"camelot-py[base]\"", "url": "https://github.com/camelot-dev/camelot", "caveats": "Digital-born PDFs only; borderless tables need tuning, and the parse-accuracy report is worth reading rather than trusting. tabula-java and pdfplumber are reasonable alternatives with different failure shapes.", "escalate": "Scanned tables, or tables whose header semantics are ambiguous, are legitimately model work — render pages with pdftoppm, give the model the image, then validate totals and row counts deterministically.", "verified": "2026-08-18" }, { "id": "office-docs-to-pdf-batch", "x": "DOCX/XLSX/PPTX (batch)", "y": "PDF", "xlabel": "office docs", "ylabel": "pdf", "tool": "libreoffice", "kind": "deterministic", "verdict": "`soffice --headless --convert-to pdf --outdir out/ *.docx` is the only free path that renders Office layout faithfully enough to hand to someone, and it batches on a server with no Office licence. Pandoc converts the content but is not a layout engine — it will not preserve a deck's slides or a spreadsheet's pagination.", "hosted": null, "local": { "tool": "libreoffice", "install": "brew install --cask libreoffice" }, "install": "brew install --cask libreoffice", "url": "https://www.libreoffice.org", "caveats": "Fidelity is high, not perfect: fonts must be installed on the host, and exotic SmartArt or macro-driven content drifts. Concurrent runs need separate profiles via `-env:UserInstallation=`.", "escalate": "Never.", "verified": "2026-08-18" }, { "id": "html-to-paginated-pdf", "x": "HTML + CSS (invoice, report, generated page)", "y": "Paginated PDF", "xlabel": "html", "ylabel": "pdf", "tool": "weasyprint", "kind": "deterministic", "verdict": "WeasyPrint implements print CSS — `@page`, running headers, page counters — so a report template paginates the same way in CI as on your laptop, with no browser in the image. Headless Chrome does the same job but its output shifts with the Chrome version, which is a bad property for anything invoice-shaped.", "hosted": null, "local": { "tool": "weasyprint", "install": "pip install weasyprint" }, "install": "pip install weasyprint", "url": "https://weasyprint.org", "caveats": "No JavaScript execution — render dynamic content to static HTML first. Very modern CSS layout features can differ from a browser's rendering.", "escalate": "Never.", "verified": "2026-08-18" }, { "id": "image-to-text", "x": "Image of text (screenshot, clean scan, photo)", "y": "Plain text", "xlabel": "image of text", "ylabel": "plain text", "tool": "tesseract", "kind": "model", "verdict": "Tesseract is a fixed local model — same input, same output, no API, nothing leaving the machine — and `--psm` is the knob that actually matters (6 for a block, 7 for a single line). Sending screenshots to a hosted vision API for text you could extract locally is a cost and privacy leak on the easy majority of cases.", "hosted": null, "local": { "tool": "tesseract", "install": "brew install tesseract tesseract-lang" }, "install": "brew install tesseract tesseract-lang", "url": "https://github.com/tesseract-ocr/tesseract", "caveats": "Wants clean, high-contrast, deskewed input; handwriting, stylized type and text over busy imagery are outside its range. When it fails it produces visible garbage, which is a useful property.", "escalate": "Handwriting, dense multi-column layouts, or text embedded in imagery — a vision model reads those. Also whenever the target is semantic (\"what is the total on this receipt\") rather than the characters themselves.", "verified": "2026-08-18" }, { "id": "speech-audio-to-transcript", "x": "Recorded speech (meeting, interview, podcast)", "y": "Transcript (text/SRT/VTT)", "xlabel": "recorded speech", "ylabel": "transcript", "tool": "whisper.cpp", "kind": "model", "verdict": "This pair has no deterministic answer; speech-to-text is model work and pretending otherwise wastes a day. whisper.cpp runs the weights locally at usable speed on Apple Silicon and writes SRT/VTT directly, so recordings never leave the machine and the marginal cost per hour of audio is zero.", "hosted": null, "local": { "tool": "whisper.cpp", "install": "brew install whisper-cpp" }, "install": "brew install whisper-cpp", "url": "https://github.com/ggml-org/whisper.cpp", "caveats": "No speaker diarization out of the box; accuracy drops with accents, crosstalk and domain jargon, and it can emit invented text over long silences. Bigger models are considerably better and considerably slower. faster-whisper is the Python/GPU alternative.", "escalate": "Already model-tier. Go further only for diarization, or for a cleanup pass on punctuation and domain terminology — and always keep the raw transcript alongside the cleaned one.", "verified": "2026-08-18" }, { "id": "messy-doc-to-schema", "x": "Messy document (invoice, contract, report PDF/DOCX)", "y": "Structured records against a schema", "xlabel": "messy document", "ylabel": "structured records", "tool": "unstructured", "kind": "hybrid", "verdict": "Split the job rather than the tool: deterministic partitioning (unstructured's parsers, or pdftotext/pdfplumber) yields elements, positions and tables, and a model is asked only for the fields that are genuinely judgment — which of these three numbers is \"the total\", is this clause an indemnity. Handing the whole PDF to an LLM and asking for JSON is the expensive failure mode: it works on the demo document and fabricates on the tenth, with no signal that it did.", "hosted": null, "local": { "tool": "unstructured", "install": "pip install \"unstructured[all-docs]\"" }, "install": "pip install \"unstructured[all-docs]\"", "url": "https://github.com/Unstructured-IO/unstructured", "caveats": "Heavy dependency tree (OCR, ONNX models) and quality varies a lot by document type; validate the extracted structure before trusting anything built on top of it.", "escalate": "The model tier is the point here — keep it narrow. Constrain output to the schema, validate every extracted number against the deterministic parse, and log both so a disagreement is visible.", "verified": "2026-08-18" }, { "id": "legacy-ebook-to-epub", "x": "MOBI/AZW3/LIT and other legacy ebook formats", "y": "EPUB", "xlabel": "legacy ebook", "ylabel": "epub", "tool": "calibre", "kind": "deterministic", "verdict": "`ebook-convert in.azw3 out.epub` carries the TOC, metadata and internal links across a dozen legacy formats — a decade of accumulated format edge cases you should not re-derive. Pandoc does not read these formats, and \"free online ebook converter\" sites mean uploading your library to a stranger.", "hosted": null, "local": { "tool": "calibre", "install": "brew install --cask calibre" }, "install": "brew install --cask calibre", "url": "https://calibre-ebook.com", "caveats": "DRM-protected files will not convert, and whether to remove DRM is your legal call, not the tool's. Heavily designed fixed-layout books reflow poorly.", "escalate": "Never.", "verified": "2026-08-18" }, { "id": "media-to-metadata-json", "x": "Photo / video / PDF file", "y": "Structured metadata (JSON)", "xlabel": "media file", "ylabel": "metadata json", "tool": "exiftool", "kind": "deterministic", "verdict": "`exiftool -json -r dir/` reads (and writes) metadata across essentially every format and tag standard, which makes it the right first move before any lossy pipeline: capture the metadata, then convert. The naive path drops timestamps and GPS during the convert step and you find out months later, with no way back.", "hosted": null, "local": { "tool": "exiftool", "install": "brew install exiftool" }, "install": "brew install exiftool", "url": "https://exiftool.org", "caveats": "Writing tags rewrites files — keep the `_original` backups, or pass `-overwrite_original` deliberately. Tag namespaces are numerous and vendor-specific.", "escalate": "Never for reading. A model is only relevant for deriving metadata that isn't there — captions, subjects, scene descriptions — which is a different job with different accuracy expectations.", "verified": "2026-08-18" }, { "id": "legacy-encoding-to-utf8", "x": "Text/CSV in a legacy encoding (CP-1252, Latin-1, Shift-JIS) showing mojibake", "y": "Clean UTF-8", "xlabel": "legacy encoding", "ylabel": "utf-8", "tool": "iconv", "kind": "deterministic", "verdict": "`iconv -f WINDOWS-1252 -t UTF-8` fixes the entire class of \"why are there  characters\" bugs in one pass once you know the source encoding; pair it with `uchardet` or `file -I` to guess. Find-and-replacing the visible mojibake by hand is the failure mode — you fix the common characters and leave the rare ones to break a parser downstream.", "hosted": null, "local": { "tool": "iconv", "install": "built-in (GNU/BSD iconv); brew install uchardet for detection" }, "install": "built-in (GNU/BSD iconv); brew install uchardet for detection", "url": "https://www.gnu.org/software/libiconv/", "caveats": "Encoding detection is a guess, not a fact — verify on a sample containing known-odd characters. `//TRANSLIT` silently degrades anything unmappable.", "escalate": "Never for the conversion; a model is at most a tie-breaker when detection is ambiguous and you have a sample of known-correct text to compare against.", "verified": "2026-08-18" } ] }