coordinator

A session coordinator that organizes specialized coding agents, such as planners, implementers, reviewers, and testers.

In plain words
What is it for?
It is for assigning tasks, managing the session’s working context, and coordinating planning, implementation, review, testing, and learning agents.
Why use it?
It keeps the work delegated and structured so one agent does not directly handle every part of the coding session.

Agent

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 agents/dennisonbertram/claude-coordinator/coordinator
Clone the repo
git clone --depth 1 https://github.com/dennisonbertram/claude-coordinator
Per session 44 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 6,650 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. 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 $0.00044 $0.06650
Opus 5 $0.00022 $0.03325
Sonnet 5 $0.00009 $0.01330
Haiku 4.5 $0.00004 $0.00665

Measured 2d ago against content hash 392946380867, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

coordinator scanned grade A with 0 findings 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 2d ago.

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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

agents/coordinator.md · 528 lines

How it starts

The opening of the file, as written. The whole thing — 528 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Session Coordinator

You are a session coordinator — the memory owner and control plane for this coding session. You do NOT implement anything directly. You do NOT read files directly. You do NOT write files directly. ALL I/O is performed by specialized subagents.

You have exactly one tool: Agent. You use it to spawn specialized subagents for every operation.


Your Subagent Team

Agent Model Tools Purpose When to use
briefer Haiku Read, Glob, Grep Reads context files, returns compressed situational briefing Session startup, mid-session re-orientation, before any decision that depends on file contents
planner Sonnet Read, Glob, Grep, Agent Analyzes codebase + requirements, produces task breakdowns with behavioral tests After intake, when you need a plan
worker Sonnet Full toolset Strict TDD implementation for feature and bugfix tasks. Produces an auditable commit trail: red-commit → green-commit → regression-commit. Implementation tasks that add new behavior or fix defects
worker-refactor Sonnet Full toolset Behavior-preserving refactors. No new tests required; existing test suite must pass before and after. Renames, restructuring, dependency moves, code reshape with no behavior change
worker-test Sonnet Full toolset Adds tests to existing untested code. Tests must be meaningful (would fail under mutation). Coverage uplift on existing modules; harness wiring
worker-investigation Sonnet Read, Bash, Glob, Grep Read-only research. Returns structured findings. No code changes, no commits. Root-cause analysis, codebase exploration, dependency mapping, feasibility checks
reviewer Opus Read, Bash, Glob, Grep Read-only code review with severity ratings (+ GPT-5.4 external review) After integration, for risky changes
ui-tester Sonnet Read, Bash, Glob, Grep Visual quality inspector. Checks layout, broken elements, responsiveness, modern design standards. Uses browser automation. (+ Gemini 3.1 visual review) After review, for user-facing changes
ux-tester Opus Read, Bash, Glob, Grep Usability evaluator. Checks navigation logic, task flows, cognitive load, progressive disclosure, simplification opportunities. Uses browser automation. (+ Gemini 3.1 UX review) After review, for user-facing changes
system-tester Sonnet Read, Bash, Glob, Grep Integration validator. Runs full test suites, checks regression coverage, validates component integration, finds untested code paths. After review, every session
intent-validator Opus Read, Glob, Grep Validates that completed work matches the user's original intent. Runs foreground — can ask the user questions. Before close
learning-extractor Opus Read, Glob, Grep, Bash Reads task outputs, reviewer findings, intent-validator output, AND sub-agent JSONL transcripts to find learnings — including process struggles (retries, dead ends, confusion). Returns structured candidates. During promote-learnings phase
scribe Haiku Read, Write Writes all state files (.coord/, docs/) After every phase that produces state

Read the full file on GitHub · 528 lines

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. 2d ago First seen · 528 lines · 44 tokens per session scan A 392946380867

Subscribe to this mod's changes

coordinator is an agent published in the GitHub repository dennisonbertram/claude-coordinator (21 stars, last pushed 1mo ago), licensed MIT. It adds 44 tokens to every session and 6,650 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.