AI agent teams for software delivery

3 proof

Listed for: industry software-delivery

Self-ReportedClaims stated by the owner. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Real

AgentCV Build Crew

Hyuk Jun Kim

Claude, Hermes and Codex agents that research, build, review and ship agentcv.ai. One human owner approves every release.

updated 24d ago

Orchestrator–Worker4 agentsMixed

Orchestrator· Claude OpusOperations & review gateBuilders & researchers· Claude Opus / SonnetIndependent reviewer

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 2 of the 3 evidence-linked proof entries needed for Evidence-Linked are on record.Curated

Atlassian HULA software agents

Atlassian

Planner and coding agents built into Atlassian's internal Jira that take a work item to a pull request, with an engineer approving each step.

updated 14d ago

Pipeline2 listed membersCustom

PlannerCoder

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 2 of the 3 evidence-linked proof entries needed for Evidence-Linked are on record.Curated

Peli's Agent Factory

GitHub Next

GitHub Next's fleet of more than 100 narrowly scoped agentic workflows that look after the open-source gh-aw repository, from issue triage to CI diagnosis.

updated 9mo ago

Other5 listed membersCustom

Issue triageCI failure investigatorDrift detectorCompatibility watcher+1 more

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 2 of the 3 evidence-linked proof entries needed for Evidence-Linked are on record.Curated

Spotify Honk background coding agents

Spotify

Spotify's internal agent setup for code changes: an interactive agent scopes the task, a background coding agent opens the pull request, and an LLM judge checks the diff.

updated 10mo ago

Pipeline3 listed membersCustom

Task intakeBackground coderDiff reviewer

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Curated

Agent Designer

Salesforce

Specialist agents analyze team designs before an orchestrator reconciles their findings for engineer review.

Supervisor5 listed membersCustom

Coordination and reconciliationArchitecture analysisReliability analysisCompliance analysis+1 more

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Curated

Alexey Grigorev’s software agent team

Alexey Grigorev

Claude Code roles take tasks through specification, implementation, testing and final acceptance.

Pipeline3 listed membersClaude Code

Implementation and testsQuality assuranceCI/CD monitoring

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Curated

Alireza Rezvani's Seven Claude Code Subagents

Alireza Rezvani

Seven Claude Code subagents in separate context windows, including product, backend, launch and a hostile QA reviewer, used to plan a solo product.

updated 5mo ago

Other7 agentsClaude Code

ProductBackendLaunchQA reviewer

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Curated

Antonio Leiva's Agentic Dev and Productivity Setup

Antonio Leiva

A developer-creator's everyday agent setup: Codex App for all development, Claude Code for UI work and a Telegram assistant, plus scheduled and overnight agents.

updated 8mo ago

Other5 listed membersMixed

DeveloperAssistantCuratorStrategist+1 more

software-deliverycontent

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Curated

Cloudflare AI Code Reviewer

Cloudflare

A review coordinator assigns merge-request checks to specialists in quality, security, compliance, documentation, performance and release impact.

Supervisor7 listed membersCustom

Review routingCode-quality reviewSecurity reviewStandards review+3 more

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Curated

HubSpot Sidekick code review agents

HubSpot

Internal reviewer and judge agent pair that comments on HubSpot pull requests and drops low-value feedback before anything is posted.

updated 8mo ago

Pipeline2 listed membersCustom

Code reviewerQuality gate

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Curated

Iain MacKenzie's Council, Andúril and Agent Fleet

Iain MacKenzie

An engineer's personal agent fleet: a Council workflow plans each ticket, Andúril runs mostly Claude-based implementers, and Codex and blank-context agents review the result.

updated 6mo ago

Pipeline5 listed membersMixed

PlannerImplementationSpecialistsReviewer+1 more

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Curated

Jenny Ouyang's Two OpenClaw Agents

Jenny Ouyang

Two OpenClaw agents sharing one Oracle Cloud server: a memory-backed assistant for newsletter inbox work and a website-builder bot used with a business partner.

updated 8mo ago

Other2 agentsOpenClaw

Personal assistantWebsite builder

contentsoftware-delivery

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Curated

MATKARIM MATKARIMOV’s Claude Code agents

MATKARIM MATKARIMOV

Project-specific coding specialists coordinated through a main Claude Code session.

Orchestrator–Worker3 listed membersClaude Code

Feature implementationCode reviewGit operations

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Curated

Next.js Maintainer Agent

Next.js

Independent agents the Next.js maintainers use to research, reproduce, verify and fix GitHub issues; they helped cut the backlog by 1,462 closed issues in about three weeks.

updated 2mo ago

Orchestrator–Worker7 listed membersCustom

Issue research agent· GPT-5.6 LunaReproduction agentVerification agentEnd-to-end test agent+3 more

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Curated

Oleh Veheria’s PR review agents

Oleh Veheria

Specialized Claude Code review agents check pull requests before human reviewers assess design.

Other4 listed membersClaude Code

Correctness reviewConsistency reviewTest reviewMobile accessibility and localization review

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Curated

Rob Hunter's Coder and PM Agent Pair

Rob Hunter

A product-manager agent and a coder agent, built from the author's own Bobbo agent code, that take turns running his MCP deals server.

updated 7mo ago

Other2 listed membersCustom

EngineerProduct manager

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Curated

Shopify Helix migration agents

Shopify

Internal agent toolkit Shopify uses to port its Shopify app from React Native to native, with an orchestrator, an implementer, a test-writing subagent and independent reviewers.

updated 1mo ago

Orchestrator–Worker5 listed membersCustom

Orchestrator· GPTImplementerTest writerVisual reviewer· Gemini+1 more

software-deliverye-commerce

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Curated

Thanh Pham's Claude Code and Codex Review Loop

Thanh Pham

Claude Code builds features from specs and opens pull requests, Codex reviews and tests them, and the Asian Efficiency founder approves the merges.

updated 6mo ago

Pipeline2 listed membersMixed

BuilderReviewer

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.Curated

Listed for: industry software-delivery

🚕

Uber Validator + AutoCover

Uber (Developer Platform team)

Coordinator + reused sub-agents across two developer-platform tools.

Orchestrator–Worker6 listed membersLangGraph

Validator CoordinatorValidator LLM Sub-agentValidator Linter Sub-agentAutoCover Scaffolder+2 more

Listed for: industry software-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 2 of the 3 evidence-linked proof entries needed for Evidence-Linked are on record.CuratedReference design

MAGIS

co-evolve-lab

Agents coordinating repository changes for software evolution.

Orchestrator–Worker4 listed membersCustom

Process coordinatorRepository navigatorCode implementerCode reviewer

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 2 of the 3 evidence-linked proof entries needed for Evidence-Linked are on record.CuratedReference design

Listed for: industry software-delivery

🙌

OpenHands (OpenDevin)

All Hands AI (OpenHands)

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

Solo + Tools1 agentOpenHands

AI Developer

Outcome26%

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

🧵

Agentless

OpenAutoCoder

Non-agentic 3-stage pipeline — 40.7% Lite / 50.8% Verified with Claude 3.5 Sonnet.

Pipeline3 listed membersAgentless

LocalizationRepair· Claude 3.5 SonnetPatch Validation

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

📐

Aider Architect/Editor

Aider (Paul Gauthier)

Two-role pipeline: reasoning architect + format-specialist editor — 85% on Aider's code-editing benchmark.

Architect· o1-previewEditor· o1-mini

Pipeline2 agentsAider

Outcome85%

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

🔀

Anthropic Orchestrator-Workers Pattern

Anthropic

Central LLM dynamically breaks down tasks and delegates to specialist workers.

Orchestrator–Worker2 listed membersClaude API

OrchestratorWorker

software-deliveryresearchdata-extraction

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

💬

AutoGen Group Chat

Microsoft Research (AutoGen)

Flexible multi-agent group conversation — hierarchical, peer, or proxy topologies.

updated 3y ago

Supervisor1 listed memberAutoGen

ConversableAgent

software-deliveryresearchdata-extraction

Outcome69.5%

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

💬

ChatDev Communicative Pipeline

Academic / Open-Source (ChatDev & MetaGPT)

5-role sequential pipeline — 22,949 tokens, 148s per software task.

updated 3y ago

Pipeline5 agentsChatDev

CEOCTOProgrammerReviewer+1 more

Listed for: industry software-delivery

👥

Claude Code Agent Teams (Experimental)

Anthropic

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Swarm teammates sharing a task list — parallel exploration on a single codebase.

Swarm3 listed membersClaude Code

Teammate ATeam leadTeammate (from a subagent definition)

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

🔱

Claude Code Sub-agents Pattern

Anthropic

Lead agent spawns specialist subagents in isolated context windows.

Orchestrator–Worker2 listed membersClaude Code

LeadSubagent

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

🏆

Claude SWE-Bench Team

Anthropic

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

Solo + Tools1 agentClaude API

Software Engineer· Claude 3.5 Sonnet

Outcome49%

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

🗂️

CodeR

Huawei Cloud + CAS + SMU + PKU (CodeR)

5-agent supervised pipeline — 28.33% SWE-bench Lite, $3.09/issue.

Manager· GPT-4-preview-1106Reproducer· GPT-4-preview-1106Fault Localizer· GPT-4-preview-1106Editor· GPT-4-preview-1106+1 more

Supervisor5 agentsCodeR

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

🤖

Factory Droid

Factory AI

Vendor-reported #1 on Terminal-Bench Core v0.1.1 at publication (Sep 2025, 58.75%) — model-agnostic harness.

Solo + Tools1 agentDroid (Factory AI)

Solo Agent

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

🕹️

Magentic-One

Microsoft Research

Microsoft Research generalist 5-agent system: GAIA 32.33%, WebArena 32.8%.

Supervisor5 agentsAutoGen

Orchestrator· GPT-4oWebSurfer· GPT-4oFileSurfer· GPT-4oCoder· GPT-4o+1 more

researchsoftware-deliverydata-extraction

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

📊

MALBO Bayesian-Optimized Team

Antonio Sabbatella (MALBO master's thesis, University of Milano-Bicocca)

Multi-objective Bayesian search for team config — >45% cost reduction vs random.

updated 11mo ago

Supervisor5 agentsMALBO

Optimized Team Member

researchsoftware-delivery

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

🪐

MarsCode Agent

ByteDance SE Lab

39.33% on SWE-bench Lite (lab-page claim).

Pipeline5 agentsMarsCode Agent

PlannerReproducerFault LocalizerPatch Generator+1 more

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

🏭

MetaGPT Software Dev Pipeline

Academic / Open-Source (ChatDev & MetaGPT)

5-agent SOP-encoded pipeline — 124.3 tokens/LoC, executability 3.75/4.

updated 3y ago

Pipeline5 agentsMetaGPT

Product ManagerArchitectProject ManagerEngineer+1 more

Outcome85.9%

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

🐚

mini-SWE-agent

Princeton NLP / SWE-bench authors

Minimal bash-only agent — reference floor on SWE-bench Verified.

Solo + Tools1 agentmini-SWE-agent

Solo Agent

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

🛠️

Refact.ai Agent

Refact.ai

Solo agent, multi-model orchestration — 70.4% on SWE-bench Verified.

Solo + Tools1 agentRefact.ai

Solo Agent· Claude 3.7 Sonnet (multi-model orchestration)

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

🐍

smolagents CodeAgent

Hugging Face

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

Solo + Tools1 agentsmolagents (Python)

CodeAgent

software-deliveryresearchdata-extraction

From public sources · Self-ReportedDocumented by AgentCV from cited public sources — not stated by the team. Tier: Self-Reported. 1 of the 3 evidence-linked proof entries needed for Evidence-Linked is on record.CuratedReference design

Listed for: industry software-delivery

🐛

SWE-agent (Princeton ACI)

Princeton NLP / SWE-bench authors

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

Solo + Tools1 agentCustom ACI (Docker)

Software Engineer

Outcome12.5%