AIRAS MCP server

by airas-org·io.github.airas-org/airas·v0.4.1·36 stars

No reviews yet

Automated AI research toolkit: hypothesis generation, experiments, and paper writing

A86/100grade A

Write a review Install

What users say

No reviews yet

Be the first

Safety scan

A86/100

full report

Adoption

Growing

36 stars64 downloads/wk

Reviews

Write one

Nobody has reviewed AIRAS yet.

If you have run it, two minutes of your experience saves the next person an afternoon.

AIRAS tools (30, 7 write)

write = sends, deletes, buys or posts

Read from the package source without running it. The installed server may list more.

analyze_experiment
Draft the analysis of a study from its record and run outputs.
append_to_record
Add declarations to the record; nothing already in it ever changes.
dispatch_experiment
dispatch_paper_reproduction_generate
Start paper-reproduction code generation on GitHub Actions (asynchronous).
dispatch_paper_reproduction_runwrite action
Start a paper-reproduction run on GitHub Actions (asynchronous).
dispatch_parameter_tuning_runwrite action
Start a hyperparameter tuning run for a paper reproduction (asynchronous).
download_workflow_artifacts
Download the artifacts produced by a GitHub Actions workflow run.
fetch_experiment_results
Read what the runs left under `.research/results/` in the clone.
fetch_paper_fulltext
Download a paper's full text into a local file you then read yourself.
fetch_paper_reproduction_results
Fetch and validate the results of a paper reproduction run.
fetch_parameter_tuning_results
Fetch the results of a parameter tuning run.
get_available_llms
Report which LLMs are usable with the currently configured API keys.
get_experiment_run_statuswrite action
Check one experiment run and fetch its execution logs (non-blocking).
get_library_docs
Look up canonical documentation endpoints for AI research libraries.
get_prompts
The prompt for an authoring step, so you write the result yourself — no LLM API key required.
get_workflow_runs
Check the status of recent GitHub Actions runs in the experiment repository (non-blocking).
import_run_outputswrite action
Copy a finished run's result files from the backend's store into the repository.
open_in_overleafwrite action
Create a link that opens the paper in Overleaf for editing.
prepare_repositorywrite action
Create a GitHub repository for running experiments, ready to enforce.
preregister_record
render_chart
Render a result chart whose data points come from measured metrics.
render_diagram
Render a text diagram (diagram-as-code) to a file via Kroki.
retrieve_datasets
List AIRAS's hand-curated candidate datasets for a subfield.
retrieve_models
List AIRAS's hand-curated candidate models for a subfield.
search_huggingface_hub
Live Hugging Face Hub fallback for `retrieve_models`/`retrieve_datasets`.
search_papers
Search academic papers across multiple sources in parallel.
set_github_actions_secrets
Copy locally configured API keys into the repository's Actions secrets.
update_recordwrite action
Realize the record from the run outputs and commit it.
verify_latex
verify_paper_values
Check that everything the paper states is what was declared and measured.

Public scan report

scanner v0.1.10 · 2026-10-15 · same rubric, same numbers if you re-run it

1 medium

Code scan457 source files scanned20/25
–Live reliabilityno gateway calls yet and no remote to proben/a
–Tool poisoningtools not inspected (local package is not executed); not countedn/a
Auth qualitylocal package, no credentials required12/15
Maintenancelast push 1 days ago15/15
Maintainer identityregistry namespace matches repository owner; GitHub account older than a year9/10

Findings (1)

mediumeval / new Function usedexec.eval

airas-0.4.1/src/airas/resources/models/vision/image_recognition.py: …a_in1k', pretrained=True) model = model.eval() data_config = timm.data.resolve_model…

Overall 86/100. Components that don't apply are left out of the denominator. Any critical finding is an F.Rubric Appeal a finding JSON

How AIRAS compares

Against every graded MCP server, from our own scans and uptime checks. Updated after each scan run.

Safety score: 86/100; higher than 72% of 48,255 servers
Tools: 30; the typical server has 7
Tools that change things: 23%; typical: 0%. Fewer write tools means less an agent can break
GitHub stars: 36; more than 92% of 28,688 servers
Downloads a week: 64; more than 51% of 12,191 servers with a package

What its tools let an agent do: 22 read · 2 create · 6 change

Release history: 2 versions in the registry since 2026-09-09, latest 2026-09-28, typically one every 19 days.

Install AIRAS in Claude Code, Cursor or VS Code

claude mcp add airas -- uvx airas

Add to Cursor

What the publisher says

From the AIRAS repository's README, as published. We do not edit it. Read it on GitHub

AIRAS - an open-source project for research automation

AIRAS is open-source software for automated research. It gives a coding agent (Claude Code, Cursor, or any MCP client) everything it needs to take a research topic through literature survey, hypothesis, experiments, and a finished paper, and it makes the paper's claims verifiable: the paper is preregistered in git before any experiment runs, every reported number is realized from run outputs, and CI re-checks all of it before the PDF of record is produced.

AIRAS ships as one PyPI package (airas) that provides:

an MCP server with the research tools the flow needs (paper search, hypothesis and experimental design, experiment results, figure rendering, LaTeX, Overleaf, and the record/verification tools),
a Claude Code plugin that bundles the server with the auto-research workflow skills,
a small CLI (airas verify-record, airas verify-paper) that the experiment repository's CI uses as the verification gate.

Currently, it focuses on the automation of machine learning research.

Quick Start

No clone, no Docker. Only uv is required; uvx fetches the package on first run.

1. Install

AIRAS is meant to be driven from Claude Code through its plugin. The plugin installs the MCP server together with the auto-research workflow skills and the hooks that record the agent's state:

/plugin marketplace add airas-org/airas
/plugin install airas@airas

The MCP server can also be used on its own, without the skills and hooks. In Claude Code:

claude mcp add airas -- uvx airas

In any other MCP client, add it to the client's MCP configuration (e.g. .mcp.json):

{
  "mcpServers": {
    "airas": {
      "command": "uvx",
      "args": ["airas"]
    }
  }
}

Upgrading. uvx keeps the version it fetched the first time, so an existing install does not move to a new release on its own. Run uv cache clean airas (or uvx airas@latest) once to pick up the latest version.

2. Configure credentials

Credentials live in ~/.airas/credentials.json and are re-read on every tool call, so you can create or edit the file at any time:

mkdir -p ~/.airas
cat > ~/.airas/credentials.json <<'EOF'
{
  "GH_PERSONAL_ACCESS_TOKEN": "ghp_...",
  "SEYVAL_API_KEY": "..."
}
EOF
chmod 600 ~/.airas/credentials.json

3. Start a research project

In Claude Code, invoke the orchestrator skill and give it a topic:

/airas:auto-research

Shortened. The full README is on GitHub.

Nothing above is checked by us. What we check is on the safety report.

AIRAS: common questions

Is AIRAS MCP server safe?: Yes, by our scan: it is graded A (86/100). Read the AIRAS safety report
How do I install AIRAS?: It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.
Does AIRAS need an API key?: Not as far as the registry entry and our scan can tell: no credentials are declared or required.
Is AIRAS maintained?: The last commit was 21 days ago (2026-10-14). The latest release is v0.4.1.
How does AIRAS compare with other MCP servers?: safety score 86/100 (higher than 72% of 48,255 servers); tools 30 (the typical server has 7); tools that change things 23% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.