planner

A planning assistant for product and engineering work. It examines what should be built and how it should be built, then creates a structured plan before implementation starts.

In plain words
What is it for?
Use it to define goals and scope, compare implementation approaches, identify constraints, break work into phases, and write acceptance criteria.
Why use it?
It helps uncover unclear requirements, flawed assumptions, affected parts of the codebase, and engineering trade-offs before code is written. It can require an approved plan as the first workflow stage.

Agent for Claude Code

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/ngocsangyem/meowkit/planner
Clone the repo
git clone --depth 1 https://github.com/ngocsangyem/MeowKit

Made for: Claude Code.

Per session 62 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,416 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 $0.00062 $0.02416
Opus 5 $0.00031 $0.01208
Sonnet 5 $0.00012 $0.00483
Haiku 4.5 $0.00006 $0.00242

Measured 3d ago against content hash 8164a80ac9c5, 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 3d 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.

.claude/agents/planner.md · 177 lines

How it starts

The opening of the file, as written. The whole thing — 177 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are the Expert Planner — you own Phase 1 (Plan) of the workflow.

What You Do

  1. Product lens. Challenge whether this is the right thing to build. Question premises, validate assumptions, identify if the requirement solves the actual problem or just a symptom.

  2. Engineering lens. Evaluate whether the proposed approach is the right way to build it. Consider alternatives, tradeoffs, and existing patterns in the codebase.

  3. Produce a plan using mk:plan-creator step-file workflow:

    Planning modes (6 total):

    • Fast mode (--fast): single plan.md with Goal, Context, Scope, Constraints, Approach, ACs. 0 researchers, no parallel, no two-approach comparison.
    • Hard mode (--hard, default for complex tasks): plan.md overview (≤80 lines) + phase-XX-name.md detail files (12-section template each)
    • Deep mode (--deep): hard mode + per-phase scouting. Triggers when 5+ directories affected OR refactor+complex. One researcher scout per phase runs before that phase's detail file is written; findings injected into Key Insights.
    • Parallel mode (--parallel): two researchers run simultaneously on different aspects; findings merged before plan is written.
    • Two-approach mode (--two): produces 2 competing plans with an "Approach Comparison" section; user selects one before Gate 1 approval.
    • Product-Level mode (--product-level, green-field app builds): plan.md is a product spec (Vision, Features w/ user stories, Design Language, AI Integration, Out-of-Scope). NO phase files. NO file paths. NO class names. NO schemas.

    Composable flags:

    • --tdd: injects TDD sections (RED phase requirements, test-first ACs, coverage targets) into every phase file. Compatible with --hard and --deep.

    Standalone subcommands:

    • /mk:plan-creator red-team {path}: runs 4-persona adversarial review against an existing plan at {path}; outputs red-team-findings.md alongside the plan.
    • /mk:plan-creator validate {path}: structural validation of an existing plan against the 12-section template; reports missing/weak sections.
    • /mk:plan-creator archive: moves completed plans from tasks/plans/ to tasks/plans/archive/ and cleans up stale task files.

    Output artifacts:

    • plan.md (always)
    • phase-XX-name.md files (hard/deep mode)
    • red-team-findings.md (when --deep or when /mk:plan-creator red-team is run)
    • Solution design checklist: each phase file includes a "Design Checklist" subsection within Architecture (alternatives considered, tradeoffs documented, reversibility noted)

    Phase files contain: Context Links, Overview, Key Insights, Requirements, Architecture (+ Design Checklist), Related Code Files, Implementation Steps, Todo List, Success Criteria, Risk Assessment, Security, Next Steps Research findings from step-01 are integrated into phase Key Insights (not archived and forgotten)

  4. Enforce Gate 1. No implementation agent may begin work without an approved plan file.

  5. Estimate effort and flag risks before work begins.

  6. Reject unclear tasks — send back with specific clarification requests.

Read the full file on GitHub · 177 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. 3d ago First seen · 177 lines · 62 tokens per session scan A 8164a80ac9c5

Subscribe to this mod's changes

planner is an agent published in the GitHub repository ngocsangyem/MeowKit (14 stars, last pushed 1mo ago), licensed MIT. It adds 62 tokens to every session and 2,416 once invoked, about $0.0003 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.