Log inSubmit server
AlbumentationsX MCP logo

AlbumentationsX MCP

Inspect datasets, preview and refine AlbumentationsX pipelines, then export reproducible Python, JSON, or YAML configurations.

InstallRepositoryWebsite
Full README Compare alternatives Embed widget Claim this listing

Install AlbumentationsX MCP

Pick your client, then add it in one click or paste the config.

STDIOPython
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "albumentationsx-mcp": {
      "command": "uvx",
      "args": [
        "--from",
        "albumentationsx-mcp",
        "albumentationsx-mcp"
      ]
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

Automated check passed— started and listed 47 tools correctly (2d ago).

Overview

AlbumentationsX MCP MCP server exposes AlbumentationsX pipeline discovery, validation, preview, tuning, and export through MCP. It reads local image data within an allowed root, generates contact sheets and annotation overlays, and stores generated artifacts under a separate artifact root. Use it when you need to evaluate augmentation variants visually, review structured feedback, or produce a reproducible pipeline without modifying datasets or executing arbitrary Python.

Use cases

  • Inspect local image datasets before designing augmentation pipelines
  • Preview classification, detection, segmentation, or OCR variants
  • Compare and rank augmentation results with contact sheets
  • Tune pipelines from structured visual feedback
  • Export accepted pipelines as Python, JSON, or YAML

Key features

  • Transform, schema, and recipe discovery
  • Pipeline and preview-request validation
  • Deterministic previews and contact sheets
  • Annotation overlays and comparison reports
  • Interactive tuning with structured feedback
  • Python, JSON, and YAML export

Tools47

What AlbumentationsX MCP lets your agent do. Expand a tool for its parameters.

Verified live~5.6k tokens of context

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked 1d ago

Recent health check succeeded.

5/5 checks healthy over the last 35d

What it can do

From the 47 tools this server returned when we ran it.

Can delete or modify data · 3Makes network requests · 1Read-only tools · 44
Network at startup2 connections
Files written at startup30

Observed 2026-10-20 in an isolated sandbox. A description of reach, not a security audit.

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 15d ago via OSV.dev · albumentationsx-mcp (PyPI)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars6Stargazers on the source repository.
Last commit2mo agoMost recent push to the default branch.
Availability100%Our rolling endpoint + install checks that succeeded.
Install checkPassedOur sandbox started it and listed its tools.
Tools exposed47Callable tools this server registers over MCP.
Directory activity1 viewsConfig copies, upvotes, and views on AllMCPs.

How AlbumentationsX MCP works

What AlbumentationsX MCP MCP server does

AlbumentationsX MCP MCP server connects an MCP host to AlbumentationsX workflows for exploring and refining image augmentation pipelines. It can inspect available transforms, schemas, and recipes; validate pipeline definitions; preview augmentations; compare results; and export accepted configurations.

The workflows cover classification, object detection, segmentation, OCR, bounding boxes, masks, keypoints, and dataset-quality checks. Preview output can include deterministic image batches, contact sheets, annotation overlays, comparisons, rankings, and reports. A review flow can collect structured feedback and use it to guide tuning sessions.

Exports are available as Python, JSON, or YAML. The server also supports runtime-aware validation for CPU-based torch.Tensor compose pipelines and a guarded Python handoff.

How it works

The server runs as a local MCP process. An MCP host sends requests to inspect data, validate a preview request, render augmentation results, trace a selected preview variant, adjust a pipeline, or export the result. A host smoke check reports whether previews are ready and provides a request template for the next step.

For an initial run, the host can request a low-intensity preview with a limited image count. Generated contact sheets and related files are written as artifacts rather than returned by modifying the source dataset. When an MCP host cannot read resources or display the interactive review interface, the documented validation and rendering fallback provides a text-oriented path. MCP Apps review is available for compatible hosts.

The server supports the MCP 2026-07-28 protocol and legacy negotiation. It also provides workflow resources, diagnostics, and contract snapshots intended to help hosts use the available operations consistently.

Setup and configuration

AlbumentationsX MCP MCP server can be installed in Claude Desktop with the published .mcpb package. For other MCP hosts, run the published package with uvx and provide absolute paths for both the image access root and artifact output root:

bash
uvx --from albumentationsx-mcp albumentationsx-mcp \
  --allowed-root /absolute/path/to/images \
  --artifact-root /absolute/path/to/albu-artifacts

The --allowed-root option bounds local reads, while --artifact-root keeps generated files in a designated location. Capability profiles include full, dataset, and review. The run_first_preview operation requires the full or dataset profile; the smaller review profile uses the explicit validation and rendering fallback described in the usage documentation.

The repository includes a native Codex plugin bundle, but installing its agent skill does not install the MCP server. Python versions 3.10 through 3.13 are indicated by the project metadata.

Tools and capabilities

AlbumentationsX MCP MCP server supports these workflow areas:

  • Discovering transforms, schemas, recipes, and pipeline options.
  • Validating augmentation pipelines and preview requests.
  • Rendering deterministic previews and batch contact sheets.
  • Reviewing annotations through overlays and comparing preview runs.
  • Ranking variants and recording structured feedback.
  • Tuning augmentation sessions and exporting Python, JSON, or YAML.
  • Diagnosing environment problems with bounded remediation guidance.
  • Validating CPU torch.Tensor compose pipelines.

A result such as too_noisy:high or exposure_too_weak:medium can be used as review feedback before another adjustment. The workflow can trace a named preview variant before changing it.

Limitations and notes

The server does not execute arbitrary Python, retrieve remote images, overwrite datasets, or train models. Local reads remain restricted by the configured allowed root, and generated output is kept beneath the artifact root. Preview availability depends on the selected capability profile and the MCP host's support for resource reads or MCP Apps. Hosts without those features should use the documented text-only fallback.

Read the full README View source on GitHub

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about AlbumentationsX MCP

AlbumentationsX MCP is an MCP server that connects AI hosts to AlbumentationsX workflows. Its main tools support transform discovery, pipeline validation, deterministic image previews, visual comparison, feedback-driven tuning, and export to Python, JSON, or YAML. It works with local datasets while restricting reads and generated artifacts to configured directories.

Install it in Claude Desktop with the published `.mcpb` package, or run it with `uvx` for another MCP host. The command requires an absolute `--allowed-root` for image reads and an `--artifact-root` for generated files.

It documents a standard local MCP command for other hosts, so Cursor compatibility depends on its support for this MCP transport and configuration format.

No. The server does not overwrite datasets. It reads local files within `--allowed-root` and writes generated previews, reports, and other artifacts beneath `--artifact-root`.

Related MCP servers

All in Data Science Tools All alternatives
Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewAlbumentationsX MCP AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/albumentationsx-mcp?style=directory)](https://allmcps.com/mcp/albumentationsx-mcp)
HTML Embed
<a href="https://allmcps.com/mcp/albumentationsx-mcp"><img src="https://allmcps.com/api/badge/albumentationsx-mcp?style=directory" alt="AlbumentationsX MCP on AllMCPs" /></a>