Pixlint

io.github.amitsingh-24Published by GitHub account amitsingh-24v1.1.0Updated Oct 24, 2026

Lint, curate & prepare computer-vision datasets from your AI assistant — MCP server, 67 tools

VerifiedSTDIODesktop onlyFiles & StorageData & AnalyticsAI & ML

Overview

AI-generated overview

PixLint is a local MCP server that lets an AI assistant lint, curate, split, augment and export computer-vision datasets.

What you can do
  • Load datasets in COCO, VOC, YOLO, KITTI or folder layouts, including cloud storage
  • Analyze quality, duplicates, integrity, class distribution, embeddings, outliers and health score
  • Run readiness diagnostics, label-error detection, natural-language queries and weak-slice discovery
  • Curate by filtering, cleaning and remapping classes into a new dataset
  • Split, augment, resize, normalize and convert formats
  • Auto-label with a pretrained detector, then export to PyTorch, TensorFlow, Ultralytics, HDF5, WebDataset, FiftyOne, CVAT, LabelMe or Hugging Face Hub, and compose multi-step pipelines
When to use it
Useful when you work with image datasets for object detection, classification or segmentation and want an assistant to diagnose problems, fix labels, clean data and produce training-ready exports conversationally instead of through a GUI or paid SaaS.
Requirements
A local Python environment with the pixlint package installed from PyPI; optional extras add torch-based features and Hugging Face export. The CV_DATA_DIR environment variable must point to the dataset directory, and CV_WORKSPACE to an output directory. Cloud storage access and Hugging Face publishing need their own credentials supplied through environment variables.
Before you install
It reads and writes files on your machine, though paths are confined to the configured data directory. Credentials for cloud storage or Hugging Face come from environment variables and should be scoped narrowly. Curation, cleaning, auto-labeling and publishing create new datasets or push data to external services such as Hugging Face Hub, so review outputs before relying on them. The README also mentions an optional HTTP mode with bearer-token authentication.

Installation

In SourceWeft

  1. Open Pixlint in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

[PixLint]

PixLint

Lint, curate, and prepare computer-vision datasets — right from your AI assistant.

[Python] [License] [MCP]

PixLint is an MCP server that gives AI assistants — Claude, Cursor, VS Code, and any MCP client — direct, conversational access to a complete computer-vision dataset toolkit: analyze quality, find duplicates and label errors, clean and curate, split, augment, convert formats, and export to every major training framework.

It runs locally over stdio, or self-hosted on the internet over authenticated HTTP.

Why PixLint

Most dataset tooling is either a paid SaaS or a heavy GUI app. PixLint is a single, open-source, self-hostable server an AI agent can drive end to end — and it does things others keep behind paid tiers:

  • 🩺 Dataset Doctor — one call runs a full diagnostic and returns a prioritized, executable fix plan.
  • Label-error detection — automatically surface images that are probably mislabeled.
  • Natural-language query — "find blurry images with a person on the left", answered over your data.
  • Weak-slice discovery — find under-represented or low-quality slices to collect or augment next.
  • Curation that writes a new dataset — clean / filter / remap, not just report.
  • Auto-labeling with a pretrained detector, and one-command Hugging Face publishing.

Features

103 operations — 67 tools, 23 resources, 13 prompts.

CategoryWhat you get
LoadCOCO · VOC · YOLO · KITTI · folder, plus cloud (S3 / GCS / Azure)
AnalyzeDuplicates · quality (blur/exposure/noise/contrast) · integrity · class distribution · embeddings · semantic search · outliers · health score
Data intelligenceDataset Doctor readiness report · label-error detection · natural-language query · weak-slice / bias discovery
CurateFilter to a subset · clean (corrupt / out-of-bounds / degenerate / duplicates) · remap classes — each produces a new dataset
Augment & transformYOLO/classification/segmentation pipelines · resize · normalize · format conversion
SplitStratified / random / temporal / grouped · k-fold · data-leakage detection
Auto-labelPretrained COCO-80 detector → pre-annotated dataset
Export & publishPyTorch · TensorFlow · Ultralytics · HDF5 · WebDataset · FiftyOne · CVAT · LabelMe · Hugging Face Hub
PipelinesCompose multi-step workflows and reuse pre-built templates

Quick Start

1. Install

bash
pip install pixlint

Optional extras add heavier capabilities:

bash
pip install "pixlint[torch]"        # embeddings, auto-labeling, label-error detectionpip install "pixlint[huggingface]"  # Hugging Face export + publishingpip install "pixlint[all]"          # everything

2. Connect your AI assistant

Claude Desktop — claude_desktop_config.json:

json
{  "mcpServers": {    "pixlint": {      "command": "pixlint",      "env": { "CV_DATA_DIR": "/path/to/your/datasets" }    }  }}

Cursor / VS Code — .cursor/mcp.json or .vscode/mcp.json:

json
{  "mcpServers": {    "pixlint": {      "command": "pixlint",      "env": { "CV_DATA_DIR": "/path/to/your/datasets" }    }  }}

CV_DATA_DIR is the directory PixLint is allowed to read datasets from.

3. Just ask

"Load my dataset at /data/coco_person, give it a readiness report, then clean it and export for YOLO."

Your assistant calls the right PixLint tools in sequence — diagnose, clean, split, export — and hands back a training-ready dataset.

Security

PixLint touches the filesystem and can be exposed to a network, so protections run on every tool call:

  • Paths are confined to your configured data directory (reads and writes).
  • Credentials come only from environment variables, never tool inputs.
  • Per-call rate limiting, concurrency limits, and audit logging.
  • Decompression-bomb protection on image decode.
  • Optional bearer-token authentication for the HTTP transport.

See the Security Guide for the full threat model and the recommended production checklist.

Documentation

GuideDescription
Getting StartedInstallation, configuration, first steps
MCP Client SetupClaude, Cursor, VS Code, and remote/HTTP hosting
API ReferenceAll 67 tools with parameters
Security GuideThreat model, configuration, hosting
Pipeline TemplatesPre-built and custom pipelines

Runnable scripts live in examples/. See CHANGELOG.md for release notes.

License

PixLint is source-available under the PolyForm Strict License 1.0.0 — see LICENSE. You may use it for permitted (noncommercial) purposes; commercial use, redistribution, or modification requires a separate license from the copyright holder. Contributions are welcome via pull request.

mcp-name: io.github.amitsingh-24/pixlint

Source: README.md at commit d9e5176

Tools

0
SourceWeft has not started this package in its sandbox yet, so its tools are not known.

Version history

1
  1. v1.1.0LatestSep 16, 2026

Score

40% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts

    Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.

  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts

    Declare a license in the repository or the registry entry.