codemie-subagents

A command that examines a project and creates tailored helper-agent files in `.claude/agents/`. The files can cover testing, architecture, code review, and cleanup.

In plain words
What is it for?
Creating or updating project-specific agents for writing tests, explaining architecture, reviewing code, and finding cleanup or refactoring work.
Why use it?
Generic helper instructions may not match a project's tools or conventions. This command adapts the helpers to the codebase and its existing guides.

Command

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 commands/codemie-ai/codemie-code/codemie-subagents
Clone the repo
git clone --depth 1 https://github.com/codemie-ai/codemie-code
Per session 26 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,729 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 $0.00026 $0.02729
Opus 5 $0.00013 $0.01365
Sonnet 5 $0.00005 $0.00546
Haiku 4.5 $0.00003 $0.00273

Measured 2d ago against content hash 945b32a0ccb9, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

codemie-subagents 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.

src/agents/plugins/claude/plugin/commands/codemie-subagents.md · 350 lines

How it starts

The opening of the file, as written. The whole thing — 350 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Codemie Subagents - Generate Project-Specific Subagent Files

Command Name: codemie-subagents Description: Generate project-specific subagent files from templates by analyzing codebase and existing guides Output: AI-optimized subagent definitions in .claude/agents/


Additional user's input

Additional context/input from user: $ARGUMENTS. Might be empty by default.

Purpose

Analyze project and generate tailored subagent files:

  • Unit Tester Agent - Project's testing patterns and framework
  • Solution Architect Agent - Project's architecture and conventions
  • Code Review Agent - Project's code standards and linting rules
  • Refactor Cleaner Agent - Project's cleanup tools and critical paths

Prerequisites

  • Project is accessible
  • Templates exist at ${CLAUDE_PLUGIN_ROOT}/claude-templates/templates/agents/
  • (Optional) Backup existing .claude/agents/ if updating agents

Safety Note: If existing agents are found, original content will be preserved during updates. However, creating a backup before running this command is recommended for recovery purposes.


🚨 SIZE LIMITS

Each generated subagent: 150-300 lines maximum

✅ Do ❌ Don't
Brief examples (10-15 lines) Extensive code blocks
Tables for patterns Long prose explanations
File:line references Full code listings
One example per pattern Multiple variations

Execution

Phase 1: Discovery

Step 1.1: Find Templates
ls ${CLAUDE_PLUGIN_ROOT}/claude-templates/templates/agents/

Expected: code-review-agent-template.md, solution-architect-agent.md, unit-tester-agent.md, refactor-cleaner-agent.md

If no templates found → Report error and stop.


Step 1.1b: Check Existing Agents
ls .claude/agents/ 2>/dev/null

Purpose: Identify existing agents to update rather than recreate

Agent Name Variations to Match:

Template Possible Existing Names
unit-tester-agent.md unit-tester.md, tester.md, unit-test-agent.md, test-agent.md
solution-architect-agent.md solution-architect.md, architect.md, architecture-agent.md
code-review-agent.md code-review.md, reviewer.md, review-agent.md, code-reviewer.md
refactor-cleaner-agent.md refactor-cleaner.md, refactor.md, cleaner.md, cleanup-agent.md

Read the full file on GitHub · 350 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 · 350 lines · 26 tokens per session scan A 945b32a0ccb9

Subscribe to this mod's changes

codemie-subagents is a command published in the GitHub repository codemie-ai/codemie-code (276 stars, last pushed 3d ago), licensed Apache-2.0. It adds 26 tokens to every session and 2,729 once invoked, about $0.0001 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.