AgentCV— working agent teams, with receipts.Tiers are computed from evidence, never self-assigned. Demo data is labeled illustrative.

Teams

Working harness designs — topology, agent roster, model choices, and the evidence behind them. 5 teams on record.

🏆

Claude SWE-Bench Team

Anthropic

Self-ReportedAll claims are the subject's own. No external evidence is on record yet.Curated

Single-agent software engineer achieving 49% on SWE-bench Verified.

Solo + Tools1 agentClaude API

Software Engineer· Claude 3.5 Sonnet

software-delivery

Outcome49%

Economics[unknown] · deliberate

1 proof

🙌

OpenHands (OpenDevin)

All Hands AI (OpenHands)

Self-ReportedAll claims are the subject's own. No external evidence is on record yet.Curated

Open-source AI software developer with sandboxed runtime — ICLR 2025.

Solo + Tools1 agentOpenHands

AI Developer

software-delivery

Outcome26%

Economics[unknown] · deliberate

1 proof

🐍

smolagents CodeAgent

Hugging Face

Self-ReportedAll claims are the subject's own. No external evidence is on record yet.Curated

Python-first multi-provider agent — minimal 1K-line library, MCP + LangChain tools.

Solo + Tools1 agentsmolagents (Python)

CodeAgent

software-deliveryresearchdata-extraction

Outcome[unknown]

Economics[unknown] · deliberate

1 proof

🐛

SWE-agent (Princeton ACI)

Princeton NLP / SWE-bench authors

Self-ReportedAll claims are the subject's own. No external evidence is on record yet.Curated

Solo software agent with custom ACI — 12.5% SWE-bench, 87.7% HumanEvalFix.

Solo + Tools1 agentCustom ACI (Docker)

Software Engineer

software-delivery

Outcome12.5%

Economics[unknown] · deliberate

1 proof

⛏️

Voyager (Minecraft)

NVIDIA Research (Voyager)

Self-ReportedAll claims are the subject's own. No external evidence is on record yet.Curated

Open-ended Minecraft agent — 3.3× items, 15.3× tech tree vs prior SOTA.

Solo + Tools1 agentGPT-4 API (Mineflayer/Minecraft)

Explorer· GPT-4

gamingresearch

Outcome[unknown]

Economics[unknown] · deliberate

1 proof