meltify

Melt files of any format into prompt-ready text where every line cites its source

Agents misread blurry digits, miss text a PDF hides, can't watch video, and paraphrase the one condition that mattered. meltify turns documents, spreadsheets, slides, mail, chat exports, archives, web pages, images, video and audio into markdown an agent can quote, puts a citation on every block, and checks answers before they go out.

Quickstart

Claude Code:

/plugin marketplace add jeon-jihyeon/meltify
/plugin install meltify@meltify

Codex, Gemini CLI, Cursor and other agents that read Agent Skills:

npx skills add jeon-jihyeon/meltify

Command line only:

uvx meltify read ./inputs

The skills call the meltify command. If it isn't installed, the plugin launcher runs it through uvx, so uv is the only thing you need.

What it does

Each command has a skill of the same name, such as meltify-ocr, that tells the agent when to run it and what to do with the result.

What read handles

Attachments, archive members and downloaded files are melted again as items of their own, up to three levels deep.

Pictures get read in the same pass. Images, scanned pages, pictures inside PDF, Word, PowerPoint, Excel and mail, and the speech and scene frames of recordings all go through local OCR and speech engines, and the text lands right where the picture was:

• Each unique picture or recording is read once, and every engine result is cached by content hash under ${XDG_CACHE_HOME:-~/.cache}/meltify, so a rerun only pays for what changed
• --budget SEC caps the time spent on uncached work (600 seconds by default). Whatever's left shows up as ocr (budget) in needs, and the next run picks up where this one stopped
• --shallow skips OCR and speech recognition and just lists what needs them. Everything else melts as usual
• When two OCR engines disagree on a number, the block ends with a > disputed line. With only one engine, it ends with > unchecked: one engine

URLs work like files. read saves a web page's main text with its heading anchors (--whole keeps the full page, --render runs it in a headless browser first), downloads documents and melts them by type, and sends video links through the media pipeline. Downloads land under meltify-out/read/web/ and are revalidated with ETags on the next run. By default it refuses private addresses, follows robots.txt and stops at 20 MB or 30 seconds. See SECURITY.md for the details.

Citations

Every result carries src and a one-line cite:

Every command prints a table by default, and the full result with --json. Exit codes:

• 0: success
• 1: violations, failed checks or other errors
• 2: usage error or bad input
• 3: a needed engine, key, binary or base module is missing

Engines

auto uses local engines only. Paid engines run only when you name them in meltify.toml or on the command line. An agent can also feed in what it reads in an image with meltify ocr --reading agent=FILE, so its own reading gets cross-checked against the local engines without any API key.

Optional parts install on request with meltify doctor --install NAME:

Recordings need ffmpeg, YouTube downloads need deno, and .ppt needs LibreOffice.

Configuration

Settings are layered, lowest precedence first:

1. The packaged defaults
2. $XDG_CONFIG_HOME/meltify/config.toml (usually ~/.config/meltify/config.toml)
3. The nearest meltify.toml, searched from the working directory up to the git root, or the file you pass with --config. Outside a git repository, only the working directory is searched
4. MELTIFY_<KEY> and MELTIFY_<SECTION>_<KEY> environment variables
5. Command-line flags

See examples/meltify.toml for a sample.

Limits

• Text drawn inside an image is just pixels, so hidden finds it only through --contrast rendering and OCR
• OCR agreement means the engines read the same value, not that the value is right. A value read only by LLM engines is never marked agreed
• Speech recognition and OCR quality depend on the engine and the input
• Linux has no Apple Vision, so with PaddleOCR alone every reading is marked unchecked
• .ppt slides need LibreOffice. Without it, the row says so in needs
• 7z and rar need the system libarchive, a separate package on Linux
• Pages and Keynote files are read only through the preview image they embed, and encrypted HWP and Numbers files aren't read at all
• The plugin launcher is a POSIX shell script, so Windows isn't supported

License

MIT. meltify depends on PyMuPDF, which is AGPL-3.0, so distributing a product that bundles it comes with AGPL obligations. The pi-heif wheels bundle libheif and libde265 (LGPL-3.0), and the iwork extra pulls in enum-tools (LGPL-3.0) through numbers-parser.