https://github.com/snakeying/textum
Structured workflow that stops AI from forgetting your requirements. 4 phases with validation gates. Not smarter AI, just controllable process. Weave ideas into code that actually works.
https://github.com/snakeying/textum
claude-code claude-skill codex codex-skill skills vibe-coding
Last synced: about 7 hours ago
JSON representation
Structured workflow that stops AI from forgetting your requirements. 4 phases with validation gates. Not smarter AI, just controllable process. Weave ideas into code that actually works.
- Host: GitHub
- URL: https://github.com/snakeying/textum
- Owner: snakeying
- License: mit
- Created: 2026-01-01T19:31:34.000Z (6 months ago)
- Default Branch: main
- Last Pushed: 2026-01-17T21:47:32.000Z (5 months ago)
- Last Synced: 2026-01-18T08:08:11.116Z (5 months ago)
- Topics: claude-code, claude-skill, codex, codex-skill, skills, vibe-coding
- Language: Python
- Homepage:
- Size: 704 KB
- Stars: 39
- Watchers: 0
- Forks: 3
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
πΈοΈ Textum
Weave your ideas into runnable code
A workflow that helps AI make fewer mistakes
Not by making AI smarter, but by making the process more controllable
English | δΈζ
> β οΈ **Note**: The old commands version is deprecated and no longer supported. Only the skill version is currently supported.
## π€ What's this all about?
Ever run into these problems?
- π¬ You spend ages explaining requirements to AI, only for it to forget what you said halfway through
- π One tiny requirement change and suddenly you're rewriting everything
- π² Every generation is a roll of the dice β quality is pure luck
**That's exactly what Textum is here to fix.**
It won't make AI any smarter, but through **structured workflow and validation**, it helps AI avoid the kind of mistakes that shouldn't happen in the first place.
---
## β¨ How does it work?
Just describe your idea in plain language. Textum breaks the whole thing down into 4 phases:
### Step 1: π Requirements Clarification (PRD)
> "So, what are we building?"
AI chats with you to turn fuzzy ideas into a clear requirements doc.
Every feature gets a number (like FEATURE-001) β no more repeating yourself later.
### Step 2: ποΈ Technical Decisions (Scaffold)
> "What's the tech stack? How do we structure this?"
Lock down the architecture, pick your technologies, define the modules.
This creates a "global context" that keeps all the coding consistent.
### Step 3: 𧩠Task Breakdown (Split)
> "What's the game plan?"
Break requirements into bite-sized tasks (Stories) β think sticky notes.
One Story, one job. Finish it, then move on.
### Step 4: π» Code Implementation (Story)
> "Let's write some code!"
AI tackles one Story at a time, laser-focused without distractions.
Each Story gets validated when done β no drifting off course.
> β οΈ **Experimental**: `Story Full Exec` supports batch execution of multiple Stories (sequential, no rollback). This feature is experimental β use at your own discretion. β οΈ
---
## π― Why all these steps?
Picture this: you hand AI a 10-page requirements doc and say "build it all."
What happens?
By feature #5, AI has already forgotten what fields were in feature #1. π
**It's not that AI is dumb β it just has limited "memory."** The more info you throw at it, the easier it loses track.
That's why Textum's philosophy is simple: **Less is more**
| The Old Way | The Textum Way |
|-------------|----------------|
| Dump all requirements on AI at once | π¦ Feed it phase by phase, only what's needed now |
| Keep saying "you know, that login API thing" | π Just say API-001 β short and unambiguous |
| One requirement change = regenerate everything | π― Only regenerate what's actually affected |
---
## π What kind of projects work best?
| Project Size | Feature Count | What to Expect | Examples |
|--------------|---------------|----------------|----------|
| π± Small | 10-15 | β
Solid results | Expense tracker, todo app, personal notes |
| πΏ Medium | 15-25 | β
Solid results | Simple blog, survey system |
| π³ Larger | 25-35 | β οΈ Needs human review | Multi-role admin panel, booking platform |
> π‘ **Even bigger?** Split it into independent sub-projects and run each through Textum separately.
**Simulation Test Reports**
SKILL version test reports:
- [1.0](./simulate-test-reports/e2e-run-1.0_EN.md)
- [1.1](./simulate-test-reports/e2e-run-1.1_EN.md)
- [1.3](./simulate-test-reports/e2e-run-1.3_EN.md)
- [1.5](./simulate-test-reports/e2e-run-1.5_EN.md)
β οΈ (Tests below are based on the old commands version using Claude Opus 4.5. Note: commands version is deprecated, for reference only)
- [V2](./simulate-test-reports/v2simulate-test-report-opus_EN.md) | [V3](./simulate-test-reports/v3simulate-test-report-opus_EN.md) | [V4](./simulate-test-reports/v4simulate-test-report-opus_EN.md)
- [V5](./simulate-test-reports/v5simulate-test-report-opus_EN.md) (complexity ramps up from V5) | [V6](./simulate-test-reports/v6simulate-test-report-opus_EN.md) | [Beta](./simulate-test-reports/beta-simulate-test-report-opus_EN.md)
### β When Textum isn't the right fit
- **Need it done yesterday**: Textum takes time to run through the workflow β not great for "hotfix and ship NOW"
- **Dead simple tasks**: Like "make this button red" β just do it, don't overthink it
- **No idea what you want**: If you can't describe it, Textum can't help π
---
## π Getting Started
### 1οΈβ£ Install uv
```bash
# macOS and Linux
curl -LsSf https://astral.sh/uv/install.sh | sh
# Windows
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"
# Or via package manager
brew install uv # macOS
pip install uv # any platform
```
> π‘ **What's uv?** A blazing-fast Python package manager (think npm, but for Python)
### 2οΈβ£ Initialize the project
In your project root:
```bash
# For Codex users
uv sync --project .codex/skills/textum/scripts
# For Claude Code users
uv sync --project .claude/skills/textum/scripts
```
β
This automatically:
- Checks your Python version (needs >= 3.11)
- Creates an isolated virtual environment (keeps your project clean)
- Installs all dependencies
### 3οΈβ£ Start building
Fire up your AI assistant (Codex/Claude Code) and say:
> "I want to start a new project with Textum"
AI will walk you through the rest π―
π **Want the technical deep-dive?** Check out [Workflow](./Workflow.md)
---
## πΈοΈ Why "Textum"?
> *"The Machine's true power was never in any single judgment, but in weaving scattered people, events, and time into a web."*
> β A tribute to *Person of Interest*
On their own, individual pieces of information mean nothing. Woven together, patterns emerge.
**Textum** is Latin for "something woven into a whole." That's exactly what this project does: it doesn't create intelligence β it weaves requirements, context, and storylines together.
Once the web is complete, the path forward reveals itself.
---
## π Found it useful?
- β **Star the repo** β Help others find it
- π **Open an Issue** β Let me know what could be better
- π¬ **Share your story** β Might turn it into a case study
Thanks! π
---
## π License
[](LICENSE)
[](https://star-history.com/#snakeying/Textum&Date)