# Accio-org/CommerceAgentBench

CommerceAgentBench: Benchmarking Long-Horizon Agents in High-Fidelity, Stateful, and Reproducible Replicas of Real Online Services

Repository: https://github.com/Accio-org/CommerceAgentBench
Canonical: https://ross.abutalabs.com/products/commerceagentbench
Homepage: https://commerce-agent-bench.site.accio.ai/
Language: HTML
License: Apache-2.0
License Family: permissive
Last push: 2026-08-24T03:20:19+00:00

## Health v2 (maintenance only)
Score: 57/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 35, longevity 2
- inputs: {"age_days": 31, "days_push": 9, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1192, forks 85 (observed 2026-08-28T04:03:56.251864+00:00)

## What it is
CommerceAgentBench is a benchmark of 107 long-horizon agent tasks evaluated in high-fidelity, stateful, reproducible replicas of real online services (Shopify, Gmail, Stripe, Jira, etc.). It ships containerized task environments with deterministic or LLM-assisted verifiers and a public leaderboard for comparing model-harness pairs.

## Use cases
- benchmark llm agents on real-world commerce workflows
- evaluate long-horizon agent task completion
- compare models on browser, CLI, and API/MCP tasks
- test agents against stateful replicas of real web services
- measure agent reliability with reproducible verifiers
- run cross-harness agent evaluations

## When to choose
- you need reproducible, stateful evaluation of agents on multi-step business workflows
- you want to compare LLMs on browser, CLI, and API/MCP task execution
- you need deterministic or LLM-assisted grading of agent outcomes

## When to avoid
- you need a lightweight single-turn QA benchmark
- your domain is unrelated to commerce or business workflows
- you want a benchmark that runs without container infrastructure

## Facets
- artifact type: dataset
- maturity: active
- function: benchmarking, agent-framework, testing, mcp
- domain: artificial-intelligence, e-commerce, large-language-models, developer-tools
- platform: python, cli
- tags: agent-evaluation, llm-benchmark, stateful-mocks, long-horizon-tasks, reproducibility, leaderboard, commerce-workflows, containerized-eval, ai-agents, docker, web-server

## Member repositories
- Accio-org/CommerceAgentBench (main) score 57

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:56.251864+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:22:26.831417+00:00, confidence not recorded.
  - readme: https://github.com/Accio-org/CommerceAgentBench (fetched 2026-08-28T04:03:56.251864+00:00, sha 65769d002167)
  - homepage: https://commerce-agent-bench.site.accio.ai/ (fetched 2026-08-29T12:29:45.682759+00:00, sha fd4e8d205cec)
- Data as of 2026-08-30T08:39:29.467469+00:00.