autoresearch-worker

autoresearch-worker is a skill for Claude Code, Codex from ObolNetwork/obol-stack. It costs 25 tokens per session (1,049 once invoked), scanned A, original, Apache-2.0.

A service that runs paid machine-learning experiments on one GPU and accepts experiment jobs over HTTP, the standard protocol used by web requests. It saves logs and results locally and provides endpoints for health, status, and results.

In plain words
What is it for?
Use it to sell GPU-backed experiment execution, receive training jobs from an experiment coordinator, run a worker on a Linux GPU host with k3s, or test the paid service locally.
Why use it?
It provides a single place to execute submitted training experiments without requiring the caller to manage the GPU. Access can be restricted so users pay before using the service.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/obolnetwork/obol-stack/autoresearch-worker
Any agent
npx skills add ObolNetwork/obol-stack --skill autoresearch-worker
Clone the repo
git clone --depth 1 https://github.com/ObolNetwork/obol-stack

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for autoresearch-worker

README.md
[![agentmods](https://agentmods.dev/badge/skills/obolnetwork/obol-stack/autoresearch-worker.svg)](https://agentmods.dev/skills/obolnetwork/obol-stack/autoresearch-worker)
Your own site
<a href="https://agentmods.dev/skills/obolnetwork/obol-stack/autoresearch-worker"><img src="https://agentmods.dev/badge/skills/obolnetwork/obol-stack/autoresearch-worker.svg" alt="Measured on agentmods" height="20"></a>
Per session 25 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,049 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00025 $0.01049
Opus 5 $0.00013 $0.00524
Sonnet 5 $0.00005 $0.00210
Haiku 4.5 $0.00003 $0.00105

Measured 3d ago against content hash 7c1260ecf8a4, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

autoresearch-worker scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

The scan reads SKILL.md. This mod also ships 1 executable file (scripts/worker_api.py), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl -s http://127.0.0.1:8080/health | jq .
internal/embed/skills/autoresearch-worker/SKILL.md · 123 lines

How it starts

The opening of the file, as written. The whole thing — 123 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Autoresearch Worker

Run a single-GPU worker that accepts train.py experiments over HTTP, executes them one at a time, stores results on disk, and exposes a simple API that can be gated with x402 using obol sell http.

When to Use

  • Selling GPU-backed autoresearch experiment execution
  • Exposing a remote POST /experiment endpoint for the autoresearch coordinator
  • Running a worker on a Linux GPU host with k3s
  • Testing the sell-side of the GPU marketplace locally before exposing it publicly

When NOT to Use

  • Coordinating experiments across many workers — use autoresearch-coordinator
  • Publishing optimized checkpoints as inference — use autoresearch
  • Selling a generic HTTP app — use sell

What the Worker Exposes

  • GET /health / GET /healthz — worker health
  • GET /status — busy/idle plus current, last, and best result
  • GET /best — best known result
  • GET /experiments/<id> — fetch a stored result
  • POST /experiment — submit a train.py experiment

The worker is intentionally simple:

  • one experiment at a time
  • one GPU worker process
  • local disk for logs/results
  • no distributed queue or scheduler inside the worker

Quick Start

1. Start the worker API

python3 scripts/worker_api.py serve \
  --repo /path/to/autoresearch \
  --data-dir /data \
  --host 0.0.0.0 \
  --port 8080 \
  --timeout 300

The repo path should point at a prepared autoresearch workdir/repo with the dependencies already available via uv.

2. Verify the worker locally

curl -s http://127.0.0.1:8080/health | jq .
curl -s http://127.0.0.1:8080/status | jq .

3. Deploy the worker behind a Kubernetes Service

For production GPU sellers, prefer k3s on the GPU host. The monetization path is cluster-based, so the worker should be reachable as a Kubernetes Service.

4. Monetize it with x402

obol sell http autoresearch-worker \
  --namespace autoresearch \
  --upstream autoresearch-worker \
  --port 8080 \
  --health-path /health \
  --wallet 0xYourWalletAddress \
  --chain base-sepolia \
  --per-hour 0.50 \
  --path /services/autoresearch-worker \
  --register-name "GPU Worker Alpha" \
  --register-description "A GPU worker for paid autoresearch experiments" \
  --register-skills devops_mlops/model_versioning \
  --register-domains research_and_development/scientific_research

Read the full file on GitHub · 123 lines

Files

What ships with it

3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 3d ago First seen · 123 lines · 25 tokens per session scan A 7c1260ecf8a4

Subscribe to this mod's changes

autoresearch-worker is a skill published in the GitHub repository ObolNetwork/obol-stack (11 stars, last pushed 4d ago), licensed Apache-2.0. It adds 25 tokens to every session and 1,049 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

cuml-machine-learning

Use for GPU-accelerated machine learning on tabular data using NVIDIA cuML. Triggers when tasks involve classification, regression, clustering, dimensionality reduction, or model training on datasets.

langchain-ai/deepagents · 43 tokens

geniml

Use Geniml for audited local genomic-interval workflows: validate BED and universe contracts, plan Region2Vec or scEmbed runs, inspect model/tokenizer compatibility, and assess consensus universes.

K-Dense-AI/scientific-agent-skills · 43 tokens

geomaster

Comprehensive geospatial science skill covering remote sensing, GIS, spatial analysis, machine learning for earth observation, and 30+ scientific domains. Supports satellite imagery processing (Sentinel, Landsat, MODIS, SAR, hyperspectral), vector and raster data operations, spatial statistics, point cloud processing…

K-Dense-AI/scientific-agent-skills · 148 tokens

digital-health-clinical-asr-finetune

Stage 4 of the Clinical ASR Flywheel. Use when priority KER is above 0.3 to run stock NeMo SFT on Parakeet TDT v2 and offline cycle N+1 re-eval. NOT for generic word boosting (use /finetune-asr).

NVIDIA/skills · 72 tokens

physicsnemo-shard-tensor

Official NVIDIA-authored guidance for PhysicsNeMo ShardTensor domain parallelism — integrate domain parallelism into training/inference scripts (new or existing) with DDP or FSDP2, write and register shard patches to enable new layers/ops, and bootstrap multi-GPU correctness tests. Use when working with ShardTensor…

NVIDIA/physicsnemo · 152 tokens

cupynumeric-migration-readiness

Pre-migration readiness assessor for porting NumPy to cuPyNumeric. Use BEFORE substantial porting work begins when the user asks whether code will scale on GPU, whether they should migrate to cuPyNumeric, which NumPy patterns transfer cleanly, what must be refactored before porting, or mentions pre-port assessment…

NVIDIA/skills · 173 tokens