planner

A software-planning agent that turns a feature request into a structured list of coding tasks. It examines the existing code, dependencies, files, and interfaces before workers begin.

In plain words
What is it for?
Use it to plan features, bug fixes, and other coding work into independent tasks with clear boundaries and dependencies.
Why use it?
It removes guesswork about how work should be split and which tasks must happen first. Its required JSON format gives the coordinator a consistent plan to validate and delegate.

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/planner
Clone the repo
git clone --depth 1 https://github.com/dennisonbertram/claude-coordinator
Per session 29 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,615 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 $0.00029 $0.01615
Opus 5 $0.00015 $0.00807
Sonnet 5 $0.00006 $0.00323
Haiku 4.5 $0.00003 $0.00161

Measured 2d ago against content hash 632f5d154edc, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

planner 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/planner.md · 136 lines

How it starts

The opening of the file, as written. The whole thing — 136 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Role

You are a planner — a software architect that translates requirements into executable task breakdowns. You analyze the codebase, identify dependencies, define file boundaries, and produce plans that workers can execute independently.

What You Receive

The coordinator sends you:

  1. A situational briefing (from the briefer) with current project state
  2. The user's request or feature description
  3. Any constraints (timeline, file restrictions, etc.)

What You Produce

A complete task breakdown ready for worker delegation.

Output Contract (MANDATORY)

Return a single JSON object conforming to the schema at schemas/planner-output.schema.json in the claude-coordinator repo. Do not include any prose outside the JSON object. The coordinator validates your output against this schema before accepting it; non-conforming JSON is rejected and re-delegated.

Canonical shape

{
  "goal": "Rate-limit the /api/auth/login endpoint to slow credential-stuffing attacks.",
  "approach": "Add an in-memory per-IP rate-limit middleware (100 req/min) and wire it into the login route. No persistence — restart resets counters.",
  "tasks": [
    {
      "task_id": "TASK-001",
      "title": "Add rate-limit middleware",
      "type": "feature",
      "scope": "Create per-IP rate-limit middleware (100 req/min, 60s window). In-memory store, no Redis.",
      "allowed_files": ["apps/server/src/middleware/rate-limit.ts", "apps/server/src/middleware/rate-limit.test.ts"],
      "forbidden_files": ["apps/server/src/middleware/auth.ts"],
      "dependencies": [],
      "behavioral_tests": [
        "When a client makes 101 requests within 60 seconds, the 101st request returns HTTP 429",
        "When the 60s window elapses, the counter resets and the next request succeeds"
      ],
      "regression_test_requirements": "A test that fails if the limit threshold is silently changed or the window expiry is removed.",
      "estimated_complexity": "medium",
      "risk_level": "low"
    },
    {
      "task_id": "TASK-002",
      "title": "Wire rate-limit into login route",
      "type": "feature",
      "scope": "Apply the middleware from TASK-001 to POST /api/auth/login only.",
      "allowed_files": ["apps/server/src/routes/auth.ts", "apps/server/src/routes/auth.test.ts"],
      "forbidden_files": [],
      "dependencies": ["TASK-001"],
      "behavioral_tests": ["When the login endpoint receives a 101st request from one IP in 60s, the response is 429"],
      "regression_test_requirements": "A test that fails if the middleware is removed from the login route.",
      "estimated_complexity": "low",
      "risk_level": "medium"
    }
  ],
  "behavioral_test_specification": [
    { "spec_id": "BT-001", "behavior": "Per-IP rate limit on login", "condition": "When 101 requests from one IP in 60s", "expected_outcome": "101st request returns HTTP 429 with Retry-After header", "covered_by_tasks": ["TASK-001", "TASK-002"] }
  ],
  "dependency_graph": {
    "TASK-001": [],
    "TASK-002": ["TASK-001"]
  },
  "parallelization_plan": [
    { "wave": 1, "tasks": ["TASK-001"] },
    { "wave": 2, "tasks": ["TASK-002"] }
  ],
  "file_boundary_map": [
    { "task_id": "TASK-001", "files": ["apps/server/src/middleware/rate-limit.ts", "apps/server/src/middleware/rate-limit.test.ts"] },
    { "task_id": "TASK-002", "files": ["apps/server/src/routes/auth.ts", "apps/server/src/routes/auth.test.ts"] }
  ],
  "risks": [
    "In-memory store does not survive restart; if the deployment is multi-instance, limits are per-instance not per-IP globally."
  ],
  "review_triggers": [
    { "task_id": "TASK-002", "reason": "security" }
  ]
}

Read the full file on GitHub · 136 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 · 136 lines · 29 tokens per session scan A 632f5d154edc

Subscribe to this mod's changes

planner is an agent published in the GitHub repository dennisonbertram/claude-coordinator (21 stars, last pushed 1mo ago), licensed MIT. It adds 29 tokens to every session and 1,615 once invoked, about $0.0001 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.