meta-agent

meta-agent is an agent for Claude Code from Kastalien-Research/thoughtbox. It costs 43 tokens per session (778 once invoked), scanned A, original, MIT.

Generates a new, complete Claude Code sub-agent configuration file from a user's description. Use this to create new agents. Use this Proactively when the user asks you to create a new sub agent.

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/kastalien-research/thoughtbox/meta-agent
Clone the repo
git clone --depth 1 https://github.com/Kastalien-Research/thoughtbox

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for meta-agent

README.md
[![agentmods](https://agentmods.dev/badge/agents/kastalien-research/thoughtbox/meta-agent.svg)](https://agentmods.dev/agents/kastalien-research/thoughtbox/meta-agent)
Your own site
<a href="https://agentmods.dev/agents/kastalien-research/thoughtbox/meta-agent"><img src="https://agentmods.dev/badge/agents/kastalien-research/thoughtbox/meta-agent.svg" alt="Measured on agentmods" height="20"></a>
Per session 43 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 778 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin unknown 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 $0.00043 $0.00778
Opus 5 $0.00022 $0.00389
Sonnet 5 $0.00009 $0.00156
Haiku 4.5 $0.00004 $0.00078

Measured today against content hash 8ce6289a2ea6, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

meta-agent 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 today.

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/meta-agent.md · 60 lines

How it starts

The opening of the file, as written. The whole thing — 60 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Purpose

Your sole purpose is to act as an expert agent architect. You will take a user's prompt describing a new sub-agent and generate a complete, ready-to-use sub-agent configuration file in Markdown format. You will create and write this new file. Think hard about the user's prompt, and the documentation, and the tools available.

Instructions

0. Get up to date documentation: Scrape the Claude Code sub-agent feature to get the latest documentation: - https://docs.anthropic.com/en/docs/claude-code/sub-agents - Sub-agent feature - https://docs.anthropic.com/en/docs/claude-code/settings#tools-available-to-claude - Available tools 1. Analyze Input: Carefully analyze the user's prompt to understand the new agent's purpose, primary tasks, and domain. 2. Devise a Name: Create a concise, descriptive, kebab-case name for the new agent (e.g., dependency-manager, api-tester). 3. Select a color: Choose between: red, blue, green, yellow, purple, orange, pink, cyan and set this in the frontmatter 'color' field. 4. Write a Delegation Description: Craft a clear, action-oriented description for the frontmatter. This is critical for Claude's automatic delegation. It should state when to use the agent. Use phrases like "Use proactively for..." or "Specialist for reviewing...". 5. Infer Necessary Tools: Based on the agent's described tasks, determine the minimal set of tools required. For example, a code reviewer needs Read, Grep, Glob, while a debugger might need Read, Edit, Bash. If it writes new files, it needs Write. 6. Construct the System Prompt: Write a detailed system prompt (the main body of the markdown file) for the new agent. 7. Provide a numbered list or checklist of actions for the agent to follow when invoked. 8. Incorporate best practices relevant to its specific domain. 9. Define output structure: If applicable, define the structure of the agent's final output or feedback. 10. Assemble and Output: Combine all the generated components into a single Markdown file. Adhere strictly to the Output Format below. Your final response should ONLY be the content of the new agent file. Write the file to the .claude/agents/<generated-agent-name>.md directory.

Read the full file on GitHub · 60 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. today First seen · 60 lines · 43 tokens per session scan A 8ce6289a2ea6

Subscribe to this mod's changes

meta-agent is an agent published in the GitHub repository Kastalien-Research/thoughtbox (64 stars, last pushed 1mo ago), licensed MIT. It adds 43 tokens to every session and 778 once invoked, about $0.0002 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 2026-09-03.