Compare

Comparing 2 teams

Row-aligned side-by-side. Highlighted cells differ across columns. [unknown] cells are honest gaps, never hidden.

Save one comparison in this browser. Not synced to your account; configurations may change.

Found your fit?

Send a setup request that references these 2 teams. Your request is saved in the admin inbox for AgentCV to review. On the next page you can choose to share your name, email and message with the owners of the teams that have one.

Request this setup

🔱

Claude Code Sub-agents Pattern

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.Reference design

🗂️

CodeR

Topology

🔱Claude Code Sub…

Orchestrator–Worker

🗂️CodeR

Supervisor

Wiring

🔱Claude Code Sub-agents Pattern

Schematic

🗂️CodeR

Real wiring

Human gates

🔱Claude Code Sub…

—

🗂️CodeR

0 gates

Agents

🔱Claude Code Sub…

[unknown]2 listed members

🗂️CodeR

5 agents

Platform

🔱Claude Code Sub…

Claude Code

🗂️CodeR

CodeR

Runs on

🔱Claude Code Sub…

Claude Code ×2

🗂️CodeR

CodeR ×5

Roster

🔱Claude Code Sub…

Claude Code Lead Agent · Lead
Claude Code
[unknown]
Claude Code Subagent · Subagent
Claude Code
[unknown]

🗂️CodeR

CodeR Manager · Manager
CodeR
GPT-4-preview-1106
CodeR Reproducer · Reproducer
CodeR
GPT-4-preview-1106
CodeR Fault Localizer · Fault Localizer
CodeR
GPT-4-preview-1106
CodeR Editor · Editor
CodeR
GPT-4-preview-1106
CodeR Verifier · Verifier
CodeR
GPT-4-preview-1106

Industries

🔱Claude Code Sub…

software-delivery

🗂️CodeR

software-delivery

Task kinds

🔱Claude Code Sub…

multi-file-codingresearchcontext-management

🗂️CodeR

github-issue-resolution

Operating since

🔱Claude Code Sub…

[unknown]

🗂️CodeR

Jun 1, 2024

Trust tier

🔱Claude Code Sub…

🗂️CodeR

Proof entries

🔱Claude Code Sub…

1 total1 with links

🗂️CodeR

1 total1 with links

Oversight

🔱Claude Code Sub…

Subagents can be constrained via tool allowlists (e.g., read-only access). Lead maintains control; subagent results are summarized back to lead context.

🗂️CodeR

None in benchmark evaluation — the Manager autonomously selects a plan and coordinates the other four agents to a single submitted patch per issue.

Source

🔱Claude Code Sub…

CuratedReference design

🗂️CodeR

CuratedReference design

Metrics

Task success rate

🔱Claude Code Sub…

[unknown]

Claude Code subagent patterns — no published benchmark figures at time of indexing.

🗂️CodeR

[unknown]

Avg cost per issue

🔱Claude Code Sub…

[unknown]

🗂️CodeR

$3.09

evidence-linkedas of Jun 1, 2024

≈299K tokens / $3.09 per issue avg; 30.39 API requests avg; hard cap $8/issue. Source: arXiv 2406.01304. [evidence_linked]

SWE-bench Lite pass@1 (single submission)

🔱Claude Code Sub…

[unknown]

🗂️CodeR

28.3%

evidence-linkedas of Jun 1, 2024

85/300 resolved (one submission per issue). Source: arXiv 2406.01304. The paper's own Results prose separately states 'addressing 84 of 300' — an internal inconsistency in the paper itself (85 elsewhere, incl. the abstract and Table 4). [evidence_linked]

Back to directory

Open Claude Code Sub-agents Pattern →Open CodeR →

Field	🔱 Claude Code Sub-agents Pattern 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	🗂️ CodeR 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
Topology	Orchestrator–Worker	Supervisor
Wiring	Schematic	Real wiring
Human gates	—	0 gates
Agents	[unknown]2 listed members	5 agents
Platform	Claude Code	CodeR
Runs on	Claude Code ×2	CodeR ×5
Roster	Claude Code Lead Agent·Lead Claude Code [unknown] Claude Code Subagent·Subagent Claude Code [unknown]	CodeR Manager·Manager CodeR GPT-4-preview-1106 CodeR Reproducer·Reproducer CodeR GPT-4-preview-1106 CodeR Fault Localizer·Fault Localizer CodeR GPT-4-preview-1106 CodeR Editor·Editor CodeR GPT-4-preview-1106 CodeR Verifier·Verifier CodeR GPT-4-preview-1106
Industries	software-delivery	software-delivery
Task kinds	multi-file-codingresearchcontext-management	github-issue-resolution
Operating since	[unknown]	Jun 1, 2024
Trust tier	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.	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.
Proof entries	1 total(1 with external links)	1 total(1 with external links)
Oversight	Subagents can be constrained via tool allowlists (e.g., read-only access). Lead maintains control; subagent results are summarized back to lead context.	None in benchmark evaluation — the Manager autonomously selects a plan and coordinates the other four agents to a single submitted patch per issue.
Source	CuratedReference design	CuratedReference design
Metrics
Task success rate	[unknown] Claude Code subagent patterns — no published benchmark figures at time of indexing.	[unknown]
Avg cost per issue	[unknown]	$3.09 evidence-linkedas of Jun 1, 2024 ≈299K tokens / $3.09 per issue avg; 30.39 API requests avg; hard cap $8/issue. Source: arXiv 2406.01304. [evidence_linked]
SWE-bench Lite pass@1 (single submission)	[unknown]	28.3% evidence-linkedas of Jun 1, 2024 85/300 resolved (one submission per issue). Source: arXiv 2406.01304. The paper's own Results prose separately states 'addressing 84 of 300' — an internal inconsistency in the paper itself (85 elsewhere, incl. the abstract and Table 4). [evidence_linked]

Back to directory

Open Claude Code Sub-agents Pattern →Open CodeR →