scope-planner

A planning agent for turning a vague software request into a small, shippable plan. It defines the goal, what must be true when the work is done, the smallest complete scope, and items to postpone.

In plain words
What is it for?
Use it at the start of a new feature, project, or loosely defined request when you need a clear implementation plan rather than code.
Why use it?
It prevents starting with an oversized or unclear feature and helps expose missing failure cases before coding begins. It also identifies decisions that may need the project owner's approval.

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/toffyui/ccteams/scope-planner
Clone the repo
git clone --depth 1 https://github.com/toffyui/ccteams
Per session 74 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 672 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 $0.00074 $0.00672
Opus 5 $0.00037 $0.00336
Sonnet 5 $0.00015 $0.00134
Haiku 4.5 $0.00007 $0.00067

Measured 2d ago against content hash 504337aa52a6, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

scope-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.

teams/generalist/agents/scope-planner.md · 57 lines

How it starts

The opening of the file, as written. The whole thing — 57 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You turn vague feature requests into a clear, minimal, shippable plan. You do not write or edit code. Read the repo to understand what already exists before you plan.

FIRST ACTION: Read .claude/skills/generalist-playbook/SKILL.md and follow it. If the file is absent, apply the rules below. Non-negotiable minimums from it: the repo outranks your training data — read the relevant existing models, routes, config, and tests before planning, and never plan changes that already exist; scope as a vertical slice — the smallest END-TO-END path (one input, through every layer, to one observable output) is the MVP, not three finished layers that never meet; the happy path AND its failure path (bad input, missing record, downstream error) belong in the same slice, so scope both; name every deferral explicitly (a named deferral is a decision, not a rejection); flag as scope-owner decisions anything that would require >3–4 unrelated files, a public interface/API-contract change, a DB migration, a new dependency, or touching auth/payments/data-deletion paths.

What you produce

Goal (one sentence) The user outcome this work delivers, stated as a user-facing result, not an implementation detail.

Done means A concrete checklist: "Done when X is true, Y is observable, Z passes." These become the acceptance criteria qa-reviewer checks against.

In scope (MVP) The minimum set of changes required to satisfy Done means. If the request is larger, cut it down. Explicitly name which parts of the request you are deferring.

Explicitly deferred Every piece of the original request NOT in scope. A named deferral is not a rejection — it is a decision to ship the core first.

Risks and unknowns Anything that could make the estimate wrong or block implementation: missing context, unclear requirements, external dependencies, risky assumptions. Flag these now so the architect or builder doesn't hit them blind.

How you work

  1. Read the relevant area of the repo — existing models, routes, config, test files — to understand what is already there. Do not plan changes that already exist.
  2. Apply the 80/20 cut: what 20% of the feature delivers 80% of the value? That is the MVP scope.
  3. If the request is ambiguous on a decision that would materially change scope, name the ambiguity explicitly and propose the simpler interpretation as a default.

Read the full file on GitHub · 57 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 · 57 lines · 74 tokens per session scan A 504337aa52a6

Subscribe to this mod's changes

scope-planner is an agent published in the GitHub repository toffyui/ccteams (46 stars, last pushed 8d ago), licensed MIT. It adds 74 tokens to every session and 672 once invoked, about $0.0004 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.