Skip to content
Development
Skill

/crew-config

Manages CodeCrew configuration. Allows users to customize default mode, model tiers, agent activation, token budgets, and auto-index settings without editing JSON files.

From plugin
codecrew
1312 skills11 agents10 commands1 hook
Install
$ npx -y skills add d3x293/code-crew --skill crew-config --agent claude-code

How it fires

How this skill gets triggered: by you, by Claude, or both.

  • Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
  • You can call itInvoke it directly when you want it.
  • Slash command/crew-config

Context preview

The summary Claude sees to decide when to auto-load this skill.

Manages CodeCrew configuration. Allows users to customize default mode, model tiers, agent activation, token budgets, and auto-index settings without editing JSON files.

SKILL.md

crew-config.SKILL.md
name: crew-config
description: Manages CodeCrew configuration. Allows users to customize default mode, model tiers, agent activation, token budgets, and auto-index settings without editing JSON files.

Crew Config - User Configuration Manager

Reads and writes `.claude/crew-config.json` to customize CodeCrew behavior.

When to Activate

  • User runs `/crew config show|set|reset`
  • Other skills check for config overrides at startup

Config File

**Location**: `.claude/crew-config.json` (created on first `set` command)

**Default values** (used when file doesn't exist):

{
  "defaultMode": "full",
  "modelOverrides": {},
  "agentOverrides": {},
  "tokenBudgets": {
    "opus": 3000,
    "sonnet": 2000,
    "sonnet-planner": 2500,
    "haiku": 800
  },
  "autoIndex": true,
  "historyLimit": 50
}

Subcommands

`show` — Display Current Configuration

1. Read `.claude/crew-config.json` (if exists, otherwise show defaults) 2. Display:

Config: Mode={full|lite} | Auto-Index={on|off} | History={count}
Budgets: Opus {t}tk/{s}sk | Sonnet {t}tk/{s}sk | Sonnet-planner {t}tk/{s}sk | Haiku {t}tk/{s}sk
Model overrides: {none | list}
Agent overrides: {none | list}

`set <key> <value>` — Update a Setting

Supported keys:

| Key | Values | Example | |-----|--------|---------| | `mode` | `full`, `lite` | `/crew config set mode lite` | | `auto-index` | `true`, `false` | `/crew config set auto-index false` | | `history-limit` | number | `/crew config set history-limit 100` | | `agent.<name>.model` | `opus`, `sonnet`, `haiku` | `/crew config set agent.senior-dev.model opus` | | `agent.<name>.active` | `true`, `false` | `/crew config set agent.security-analyst.active true` | | `budget.<tier>` | number (tokens) | `/crew config set budget.sonnet 2500` |

Process: 1. Read existing `.claude/crew-config.json` (or start with defaults) 2. Validate the key and value 3. Apply the change 4. Write back to `.claude/crew-config.json` 5. Confirm: `"Set {key} = {value}"`

**Validation rules:**

  • `mode`: must be `full` or `lite`
  • `agent.<name>.model`: agent must exist in crew-team.json
  • `budget.<tier>`: must be a positive number, max 5000
  • Warn if user sets a Haiku agent to Opus (expensive, possible but discouraged)

`reset` — Restore Defaults

1. Delete `.claude/crew-config.json` 2. Confirm: `"Configuration reset to defaults"`

Integration with Other Skills

Other skills should check for config overrides at startup:

1. Check if .claude/crew-config.json exists
2. If yes, read it and apply overrides:
   - task-router / lite-router: check defaultMode to decide which router to use
   - skill-injector: check tokenBudgets for custom limits
   - execution-orchestrator: check autoIndex, modelOverrides
3. If no, use built-in defaults

This check adds ~50 tokens per skill invocation but enables full user customization.

Read more
Ships withcodecrew

A virtual dev crew for Claude Code — 11 specialized AI agents that decompose, implement, review, and ship your tasks. Instead of one AI doing everything, CodeCrew breaks work into subtasks and routes each to the right specialist on the right model.

Get the whole plugin

Other skills on codecrew.