AlbumentationsX MCP
AlbumentationsX MCP is a Model Context Protocol server that helps computer vision practitioners discover, build, validate, preview, tune, and export image augmentation pipelines using AlbumentationsX.
Transform Discovery & Inspection
Search transforms by query, target type, bbox type, or transform type
Retrieve detailed parameter schemas, target support, and summaries for specific transforms
Pipeline Management
Recommend starter pipelines for CV tasks (classification, detection, segmentation, OCR) at low/medium/high intensity
Recommend full recipes including quality profiles and preview workflows
Validate, explain, and adjust pipelines based on structured feedback tags
Export validated pipelines as Python, JSON, or YAML
Environment & Diagnostics
Diagnose MCP setup, filesystem/root access, and artifact write permissions
Run read-only preflight (smoke) checks to ensure readiness for local preview rendering
Preview Rendering
Plan dataset onboarding and validate preview requests before rendering
Render deterministic single-image or batch previews and contact sheets
Preview Comparison & Analysis
Compare two preview runs side-by-side with quality summaries
Rank multiple candidate pipelines against a baseline
Score dataset-level preview candidates as decision sets
Tuning Sessions
Start persistent, multi-step tuning sessions to iteratively refine pipelines
Record steps, summarize progress, list, export, close, archive, and clean up sessions
Feedback & Decisions
Record and list granular per-image preview feedback (e.g.,
too_noisy:high)Persist and list local tuning decisions
List all accepted feedback tags and task-aware quality profiles
Reporting & Export
Export tuning reports (Markdown/JSON) and visual preview reports (Markdown/HTML) with ranking, contact sheets, and decisions
Preview Run Management
List, retrieve manifests for, delete, and clean up older preview runs and their artifacts
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@AlbumentationsX MCPrecommend a pipeline for object detection"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
AlbumentationsX MCP
Model Context Protocol server for AlbumentationsX: inspect datasets, preview augmentations, refine them with visual feedback, and export reproducible pipelines.
Ask an MCP host for several robustness variants, reject an excessive result such as too_noisy:high, compare the adjusted batch previews, and export the accepted pipeline.
Install
Claude Desktop
Download the latest albumentationsx-mcp.mcpb, install it from Settings -> Extensions -> Advanced settings, and select separate image and artifact directories.
Other MCP Hosts
Run the published server with bounded local access:
uvx --from albumentationsx-mcp albumentationsx-mcp \
--allowed-root /absolute/path/to/images \
--artifact-root /absolute/path/to/albu-artifactsCopyable Claude Code, Cursor, Codex, and JSON configurations are in the install guide. The repository also contains a native Codex plugin bundle. npx skills add dKosarevsky/albu-mcp installs agent guidance, not the MCP server.
Related MCP server: ZenML MCP Server
First Preview
After connecting the server, ask your host:
Use AlbumentationsX MCP on /absolute/path/to/images.
Run the smoke check, start with a low-intensity pipeline, validate the request,
render one variant per image, and show me the contact sheet before exporting anything.When the host exposes resource reads, read
albumentationsx://examples/client-smoke; otherwise callrun_host_smoke_checkdirectly.Continue only when
preview_readyis true, then use the returnedpreview_request_template.Call
validate_preview_requestbefore rendering and compare preview runs before accepting a candidate.Give concrete feedback such as
too_noisy:highorexposure_too_weak:medium, then export the final pipeline.
If setup fails, read albumentationsx://diagnostics/guide and call diagnose_environment for bounded remediation actions.
Capabilities
Transform discovery, schemas, recipes, and pipeline validation.
Classification, detection, segmentation, OCR, bbox, mask, keypoint, and dataset-quality workflows.
Deterministic previews, contact sheets, annotation overlays, comparison, ranking, and reports.
Interactive MCP Apps review with a text-only fallback for other hosts.
Structured feedback, tuning sessions, and Python, JSON, or YAML export.
Stable agent workflow resources, prompts, diagnostics, and reviewed contract snapshots.
The server does not execute arbitrary Python, fetch remote images, overwrite datasets, or train models. Reads are restricted by --allowed-root; generated files stay under --artifact-root.
Integrations
Documentation
server.json: public MCP Registry metadata.
Development
uv sync --all-extras --dev
uv run pytest
uv run ruff check .
uv run ruff format --check .
uv run ty checkLicensed under AGPL-3.0-or-later.
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Latest Blog Posts
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/dKosarevsky/albu-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server