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/jimmyma99/treeskill/claude-mdgit clone --depth 1 https://github.com/JimmyMa99/TreeSkillWrote 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.
[](https://agentmods.dev/instructions/jimmyma99/treeskill/claude-md)<a href="https://agentmods.dev/instructions/jimmyma99/treeskill/claude-md"><img src="https://agentmods.dev/badge/instructions/jimmyma99/treeskill/claude-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.01248 | $0.01248 |
| Opus 5 | $0.00624 | $0.00624 |
| Sonnet 5 | $0.00250 | $0.00250 |
| Haiku 4.5 | $0.00125 | $0.00125 |
Grade A, and why
TreeSkill CLAUDE.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 3d 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 — 146 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
Project Overview
TreeSkill is a train-free prompt optimization framework using Textual Gradient Descent (TGD). It treats system prompts as "weights" and interaction history as "dataset" to iteratively improve prompts via API calls only—no model training required.
Commands
# Install
pip install -e .
# Run tests
pytest test_*.py
# Start interactive CLI with skill
python -m treeskill.main --skill <path-to-skill.yaml-or-directory>
# Start with config file
python -m treeskill.main --config <config.yaml> --skill <skill>
# Run demos
python demo/demo_from_scratch.py # Build skill from scratch
python demo/demo_from_skill.py # Optimize existing skill
# Example scripts
python example_optimizer.py # Interactive optimization demo
python example_tools.py # Tool registration demo
python example_fully_automatic.py # Fully automated optimization
python example_tree_optimization.py # Tree-aware optimization (split/prune)
python example_load_skill_and_config.py # Load stored skill + config
Architecture
Core Layers
treeskill/
├── core/ # Core abstraction layer (v0.2)
│ ├── abc.py # Abstract base classes (OptimizablePrompt, ModelAdapter, etc.)
│ ├── optimizer.py # TrainFreeOptimizer - TGD algorithm
│ ├── tree_optimizer.py # TreeAwareOptimizer - auto split/prune
│ ├── strategies.py # Optimization strategies (conservative/aggressive/adaptive)
│ └── validators.py # Prompt validators
│
├── adapters/ # Model adapters
│ ├── openai.py # OpenAI-compatible APIs
│ └── anthropic.py # Anthropic Claude
│
├── registry.py # Plugin registry (@adapter, @optimizer, @hook)
├── tools.py # Tool registry (Python functions, HTTP, MCP)
│
└── [legacy] # v0.1 backward compat
├── schema.py, skill.py, skill_tree.py
├── checkpoint.py, storage.py
└── optimizer.py (APOEngine)
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.
- 3d ago First seen · 146 lines · 1,248 tokens per session scan A 991f31a34e27
TreeSkill CLAUDE.md is an instructions file published in the GitHub repository JimmyMa99/TreeSkill (23 stars, last pushed 5mo ago), licensed Apache-2.0. It adds 1,248 tokens to every session, about $0.0062 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
Awesome-Prompt-Engineering AGENTS.md
Instructions for natnew/Awesome-Prompt-Engineering, covering agents.md, repository north star, agent role, trust boundary and read order.
SkillOpt AGENTS.md
Instructions for mitkox/SkillOpt, covering agent instructions for skillopt, project identity, default example workflow, documentation expectations and repo hygiene.
vscode-copilot-chat model-prompts.instructions.md
Model-specific prompt authoring and registry guidelines.
comfy-prompt-studio AGENTS.md
Instructions for yxhpy/comfy-prompt-studio, covering agents.md - ai 代理配置文档, ai 提供商, 1. ollama (默认), 2. gemini and 提示词生成服务.
ZipAI CLAUDE.md
Claude Code instructions for nickdesi/ZipAI, covering claude.md — zipai: ultra-dense token optimizer, rules, 1. zero filler, 2. ambiguity and 3. prompt caching.
ultra-max-token-burner AGENTS.md
Instructions for floytra-dev/ultra-max-token-burner: When the user asks to expand, operationalize, professionalize, govern, validate, or make a prompt more manager-visible, use the Ultra Max Token Burner workflow in SKILL.md.