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.
npx agentmods add instructions/backspace-shmackspace/claude-devkit/gemini-mdgit clone --depth 1 https://github.com/backspace-shmackspace/claude-devkitWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00729 | $0.00729 |
| Opus 5 | $0.00365 | $0.00365 |
| Sonnet 5 | $0.00146 | $0.00146 |
| Haiku 4.5 | $0.00073 | $0.00073 |
Grade A, and why
claude-devkit GEMINI.md 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.
How it starts
The opening of the file, as written. The whole thing — 84 lines — stays where its author put it; the contents beside it link to each section on GitHub.
GEMINI.md
Project Overview
This project, claude-devkit, is a complete development toolkit for creating and managing skills for the Claude Code AI agent. The "code" in this project primarily consists of Markdown files that define "skills"—structured workflows that the Claude Code agent can execute.
The project provides:
- Skills: Pre-built workflows for common development tasks like planning (
/architect), implementation (/ship), security audits (/audit), and documentation synchronization (/sync). - Generators: Python scripts to scaffold new skills and agents from templates.
- Templates: A set of standard templates for different types of skills and agents.
- Configuration: JSON files that define patterns for skills and tech stacks.
- Deployment Scripts: Shell scripts to install and deploy the skills to the Claude Code environment.
The core of the project is the skills directory, where each subdirectory represents a skill and contains a SKILL.md file. This file defines the skill's workflow, including its steps, the tools to use, and the logic for handling different scenarios.
Building and Running
This project does not have a traditional build process. Instead, it uses a set of scripts to generate, validate, and deploy skills.
Installation
To set up the claude-devkit, run the installation script:
./scripts/install.sh
This script sets up the necessary environment variables and aliases to use the generators and other tools.
Deploying Skills
To deploy the skills to the Claude Code environment, run the deployment script:
./scripts/deploy.sh
You can also deploy a specific skill by providing its name as an argument:
./scripts/deploy.sh <skill-name>
Running Tests
The project includes a test suite for the skill generator. To run the tests, execute the following command:
bash generators/test_skill_generator.sh
Development Conventions
Creating a New Skill
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.
- 2d ago First seen · 84 lines · 729 tokens per session scan A 4a8f8afc6941
claude-devkit GEMINI.md is an instructions file published in the GitHub repository backspace-shmackspace/claude-devkit (15 stars, last pushed 8d ago), licensed MIT. It adds 729 tokens to every session, about $0.0036 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.
Other instructions, from other repositories
nextclaw AGENTS.md
Instructions for Peiiii/nextclaw, covering nextclaw ai 常驻内核, 产品愿景, 沟通与推进, 深思与目标模式 and 协作与 git 安全.
slipway AGENTS.md
Instructions for signalridge/slipway, covering slipway agent principles, user owns the process, investigate before clarifying, report, do not certify and maintain the whole surface.
lead-dev-os CLAUDE.md
Instructions for CaptainMe-AI/lead-dev-os, covering claude.md, project, terminology, plugin architecture and repository structure.
agentic-rd-skill AGENTS.md
AGENTS.md instructions for TheStreamCode/agentic-rd-skill, covering repository instructions, purpose and sources of truth, repository invariants, implementation style and documentation synchronization.
skills-driven-development CLAUDE.md
Instructions for zakelfassi/skills-driven-development, covering skdd — agent instructions, active goal, house rules, non-goals (binding, from roadmap.md) and operational guardrails.
ccpr CLAUDE.md
Claude Code instructions for jonase47/ccpr, covering global preferences (local, interactive mode), language & communication, personal context, infrastructure and code standards.