# Rapid Domain Mastery
**A portable Agent Skill for deep, multi-source learning across major AI coding agents.**
[](LICENSE)
[](https://skills.sh)
[](#validation)
---
## What It Does
Rapid Domain Mastery is a structured reasoning workflow for tasks where a user needs to **learn an unfamiliar field quickly from many sources**. It goes far beyond plain summarization by executing a multi-phase analytical pipeline.
| Phase | What It Produces |
| :--- | :--- |
| Corpus Audit | Quality assessment of the source materials, identifying gaps and biases |
| Field Cartography | A structured map of the domain's key concepts, boundaries, and relationships |
| Mental Model Extraction | The 3-5 core expert mental models that practitioners actually use |
| Debate Cartography | A map of genuine disagreements between experts, schools, or methods |
| Assumption Mining | Hidden assumptions and unstated prerequisites that sources take for granted |
| Prerequisite Ladder | An ordered sequence of concepts the learner must master first |
| Diagnostic Questions | Oral-exam-style questions that test for real understanding, not recall |
| Tutor Loop | An interactive drilling mode with Socratic correction |
| Sprint Planner | A time-budgeted study plan when the user has a deadline |
Every claim is tagged with epistemic labels (`[Consensus]`, `[Debate]`, `[Open]`, `[Gap]`, `[Prereq]`, `[Inference]`) to prevent hallucinated confidence.
## Supported Agents
This repository ships with native, compatibility, and manual adapters for the following agents and platforms. Support levels are deliberately tiered:
- **native**: the repo includes a host-specific skill path or documented first-class mechanism.
- **native-compatible**: the repo includes a compatible path or wrapper, but host behavior should be smoke-tested before broad claims.
- **native-manual-install**: the host supports the workflow concept, but users should import or install it through that host's documented flow.
- **manual**: the repo provides a prompt/conventions adapter, not native Agent Skills loading.
- **directory-listed**: listed by the Agent Skills ecosystem, but not independently vendor-verified here.
| Agent | Support Level | Mechanism |
| :--- | :--- | :--- |
| OpenAI Codex | native | `.agents/skills/`, `AGENTS.md` |
| Claude Code | native | `.claude/skills/`, `CLAUDE.md` |
| GitHub Copilot | native | `.github/skills/`, `copilot-instructions.md` |
| Cursor | native-compatible | `.cursor/rules/`, `.claude/skills/` |
| Kiro | native | `.kiro/skills/`, `.kiro/steering/` |
| Cline | native | `.cline/skills/`, `.clinerules/` |
| Roo Code | native | `.roo/skills/` |
| Augment | native | `.augment/skills/`, `AGENTS.md` |
| OpenCode | native | `.opencode/skills/`, `AGENTS.md` |
| Goose | native | `.goose/skills/`, `.goosehints` |
| Factory / Droid | native | `.factory/skills/` |
| Windsurf | native-manual-install | `.windsurf/rules/`, `AGENTS.md` |
| Antigravity | native-manual-install | `.agent/rules/`, `GEMINI.md` |
| Gemini CLI | native-manual-install | `GEMINI.md`, `AGENTS.md` |
| OpenHands | native-different-format | `adapters/openhands/` |
| Amazon Q (legacy) | manual-legacy | `.amazonq/rules/` |
| Continue | manual | `adapters/continue/` |
| Aider | manual | `adapters/aider/` |
Additional agents listed by the [Agent Skills directory](https://skills.sh) (AMP, ClawdBot, Kilo, Trae, VS Code) are ecosystem compatibility signals, not file-path-level verification. See [`VERIFIED-SUPPORT-MATRIX.md`](VERIFIED-SUPPORT-MATRIX.md) and [`support/evidence.md`](support/evidence.md) for the full breakdown.
## Quick Start
Do not clone this whole repository into a generic `.skills/` folder. The host needs the skill files at the path it actually loads.
**Option 1 โ Codex / AGENTS-compatible projects:**
```bash
git clone https://github.com/Emily2040/rapid-domain-mastery.git /tmp/rapid-domain-mastery
mkdir -p .agents/skills
cp -R /tmp/rapid-domain-mastery/.agents/skills/rapid-domain-mastery .agents/skills/
cp /tmp/rapid-domain-mastery/AGENTS.md .
```
**Option 2 โ Claude Code:**
```bash
git clone https://github.com/Emily2040/rapid-domain-mastery.git /tmp/rapid-domain-mastery
mkdir -p .claude/skills
cp -R /tmp/rapid-domain-mastery/.claude/skills/rapid-domain-mastery .claude/skills/
cp /tmp/rapid-domain-mastery/CLAUDE.md .
```
**Option 3 โ GitHub Copilot coding agent:**
```bash
git clone https://github.com/Emily2040/rapid-domain-mastery.git /tmp/rapid-domain-mastery
mkdir -p .github/skills
cp -R /tmp/rapid-domain-mastery/.github/skills/rapid-domain-mastery .github/skills/
cp /tmp/rapid-domain-mastery/.github/copilot-instructions.md .github/
```
**Option 4 โ Install via skills.sh where the package is listed and your client supports it:**
```bash
skills install rapid-domain-mastery
```
**Option 5 โ Copy the files you need manually:**
1. Copy `AGENTS.md` and `.agents/skills/rapid-domain-mastery/` into your project root.
2. Add the thin wrapper for your agent (e.g., `CLAUDE.md`, `.cursor/rules/`, `.kiro/steering/`).
3. Done.
## Example Prompt
```text
Use rapid-domain-mastery on this corpus of papers, docs, and notes.
I need a field map, the 3-5 core expert mental models, real debates,
hidden assumptions, prerequisite ladder, diagnostic questions, and a
two-week study sprint.
```
## Repository Structure
```
.
โโโ AGENTS.md # Universal persistent guidance
โโโ CLAUDE.md # Claude Code thin wrapper
โโโ GEMINI.md # Gemini CLI thin wrapper
โโโ LICENSE # MIT License
โโโ README.md # This file
โ
โโโ .agents/skills/rapid-domain-mastery/ # Canonical skill source
โ โโโ SKILL.md # Skill entry point (< 500 chars)
โ โโโ references/ # Progressive-disclosure modules
โ โ โโโ 00-orchestrator.md # Main workflow router
โ โ โโโ 00-routing-and-boundaries.md # When to use / not use
โ โ โโโ 10-input-contract.md # Input normalization
โ โ โโโ 20-corpus-audit.md # Source quality assessment
โ โ โโโ 30-field-cartography.md # Domain mapping
โ โ โโโ 40-mental-model-extraction.md # Expert model extraction
โ โ โโโ 50-debate-cartography.md # Disagreement mapping
โ โ โโโ 60-assumption-mining.md # Hidden assumption detection
โ โ โโโ 70-prerequisite-ladder.md # Learning sequence
โ โ โโโ 80-diagnostic-question-design.md
โ โ โโโ 90-tutor-loop.md # Interactive drilling
โ โ โโโ 100-study-sprint-planner.md
โ โ โโโ 110-application-and-transfer.md
โ โ โโโ 120-domain-adapters.md
โ โ โโโ 130-mcp-and-tooling.md
โ โ โโโ 140-output-templates.md
โ โ โโโ 150-quality-gates.md
โ โ โโโ 160-failure-modes.md
โ โโโ templates/ # Structured output templates
โ โโโ scripts/ # Skill-local helper scripts
โ
โโโ .claude/ .cursor/ .kiro/ ... # Agent-specific mirrors & wrappers
โโโ adapters/ # Manual adapters (Aider, Continue, etc.)
โโโ docs/assets/ # README and documentation images
โโโ .github/workflows/ # Pack validation CI
โโโ interop/ # A2A / ACP / MCP interop notes
โโโ scripts/ # Repo-level maintenance scripts
โ โโโ validate_pack.py # Mirror parity + structural checks
โ โโโ sync_adapters.py # Sync canonical โ all mirrors
โ โโโ check_local_links.py # Markdown local-link checker
โ โโโ corpus_inventory.py # Inventory helper
โโโ support/ # Verification metadata
โโโ support-matrix.json
โโโ support-matrix.csv
โโโ verification-notes.md
```
## Validation
Run the built-in validation to confirm all mirrors are in sync and structural rules are met:
```bash
python3 scripts/validate_pack.py
python3 scripts/check_local_links.py
```
To re-sync all agent mirrors from the canonical source after making changes:
```bash
python3 scripts/sync_adapters.py --dry-run
python3 scripts/sync_adapters.py
python3 scripts/validate_pack.py # always validate after sync
```
## Key Documentation
| Document | Purpose |
| :--- | :--- |
| [`VERIFIED-SUPPORT-MATRIX.md`](VERIFIED-SUPPORT-MATRIX.md) | Full agent support matrix with verification sources |
| [`INSTALL-BY-AGENT.md`](INSTALL-BY-AGENT.md) | Per-agent installation instructions |
| [`support/evidence.md`](support/evidence.md) | Public evidence ledger for compatibility claims |
| [`STANDARDS.md`](STANDARDS.md) | Design standards and conventions used |
| [`MCP-RECOMMENDATIONS.md`](MCP-RECOMMENDATIONS.md) | Recommended MCP servers for enhanced workflows |
| [`interop/A2A-ACP-MCP.md`](interop/A2A-ACP-MCP.md) | Agent-to-Agent and protocol interop notes |
## Maintenance
When updating the skill workflow:
1. Edit only the canonical source in `.agents/skills/rapid-domain-mastery/`.
2. Run `python3 scripts/sync_adapters.py` to propagate changes.
3. Run `python3 scripts/validate_pack.py` to verify.
4. Update `support/support-matrix.json` if agent compatibility changes.
When vendor documentation changes, update the relevant wrapper and the support matrix. Keep the canonical skill semantics stable; change only wrappers and install notes unless the reasoning workflow itself needs improvement.
## Author
Created by **Iamemily2050**
- [GitHub](https://github.com/Emily2040)
- [Website](https://Iamemily2050.com)
- [X / Twitter](https://x.com/iamemily2050)
- [Instagram](https://www.instagram.com/iamemily2050)
## License
[MIT](LICENSE) โ Copyright (c) 2026 Iamemily2050