skills.rest
ek33450505
ek33450505COMMUNITY
Explore all skills by this author →
WebsiteGitHub

thinking-budget

Manage per-agent thinking budget JSON configurations for Anthropic agent teams.

8|1|Updated Mar 20, 2026
Search Tags:#configuration#anthropic#agent-config#token-budget#agent-roles#thinking-budget
One-click install
npx skills add https://github.com/ek33450505/claude-agent-team --skill thinking-budget
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: thinking-budget
Source: https://github.com/ek33450505/claude-agent-team/tree/main/skills/thinking-budget
Command: npx skills add https://github.com/ek33450505/claude-agent-team --skill thinking-budget

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Centralizes and documents per-agent thinking budget configuration so teams can control how much chain-of-thought reasoning each agent is allowed, avoid zero-budget errors for unspecified agents, and allocate token budget according to role importance.

Core Features & Use Cases

  • Centralized JSON configuration stored at ~/.claude/config/thinking-budgets.json for system-wide budget defaults.
  • Tiered budgets with recommended values for common roles: 0 for mechanical reviewers, 1024 as a safe default, 2048 for light reasoning, 4096 for moderate analysis, and 8192 for deep reasoning roles.
  • Per-agent override via the agent frontmatter thinking_budget field to customize budgets for planners, orchestrators, code-writers, debuggers, and reviewers.
  • Use case: increase the planner and orchestrator budgets for multi-file planning tasks while keeping haiku reviewers at zero to save tokens.

Quick Start

Update the agent's thinking_budget frontmatter or edit the JSON file at ~/.claude/config/thinking-budgets.json to set the numeric token budget for the target agents.

Frequently Asked Questions about thinking-budget

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I configure thinking budgets for Anthropic agent teams?▼

Configure thinking budgets by editing the JSON file at ~/.claude/config/thinking-budgets.json or setting the thinking_budget frontmatter field to assign numeric token limits for each agent role.

What thinking budget values should I set for different agent roles?▼

Recommended thinking budget values are 0 for mechanical reviewers, 1024 as a safe default, 2048 for light reasoning, 4096 for moderate analysis, and 8192 for deep reasoning roles like planners and orchestrators.

How do I prevent zero-budget errors for unspecified agents in my local agent swarm?▼

Prevent zero-budget errors by centralizing default configurations in the ~/.claude/config/thinking-budgets.json file, which ensures all CAST-managed local agents have a valid token budget allocated.

Can I override the default thinking budget for a specific agent?▼

Yes, you can override default thinking budgets for specific agents by adding the thinking_budget frontmatter field to the agent configuration to customize token limits for planners, code-writers, debuggers, or reviewers.

Does the thinking-budget configuration work with CAST-managed local agent swarms?▼

Yes, thinking-budget configuration applies directly to CAST-managed local agent swarms, allowing you to control chain-of-thought token usage across different roles like reviewers, code-writers, debuggers, planners, and orchestrators.

Why allocate different token budgets to agents in an Anthropic agent team?▼

Allocating different token budgets allows you to save tokens by keeping mechanical reviewers at zero while increasing reasoning depth for planners and orchestrators handling multi-file planning tasks.

Related Skills

Explore More Software Engineering Skills→
monoesmonoes
15

swarm-orchestration

Orchestrate multi-agent swarms for parallel task execution in Monomind workflows.

Official
Advanced
engineers-hub-ltd-in-house-projectengineers-hub-ltd-in-house-project

alerting-slo

Defines SLIs, SLOs, and multi-tiered alerts based on error budget consumption.

Official
Advanced
QredenceQredence
108

dspy-agent-framework-integration

Integrate DSPy with the Microsoft Agent Framework for typed signatures and orchestration.

Official
Advanced
saeed-vayghansaeed-vayghan
30

prompt-engineer

Design, evaluate, and manage prompt architectures for large language models.

Community
Advanced