✨ What NeverMIN does
| Feature |
Description |
| 🧭 Entry points & hubs |
Heuristic centrality score highlights where to start reading |
| 🔀 Main flow diagram |
Mermaid flowchart of Input → Process → Output |
| 🗺️ Learning Mind Map |
Suggested exploration order for onboarding |
| 💬 Explain Code |
Optional LLM narration — local (Ollama) or cloud |
| 📜 Git History insight |
Alive/frozen code, ownership, hidden file coupling |
Insights are based on heuristic graph analysis (degree + centrality) — a great onboarding map, not a substitute for a full architecture review.
Repo: github.com/abdulwahidrukua/NeverMIN
🔒 Privacy & Security
NeverMIN is built so you can map a private or proprietary codebase without sending source to a third party — as long as you stay on the local path.
The sidebar asks you to choose before other features unlock. Change it any time via Settings → Change privacy mode (NeverMIN: Choose Private / Public Codebase).
| Mode |
Behavior |
| 🔐 Private codebase |
Local-only: parse/graph/git stay on disk · LLM locked to Ollama · cloud API keys cleared |
| 🌐 Public codebase |
Same local tools, plus optional cloud LLM (Gemini, OpenAI, …) once you save a key |
What never leaves your machine (either mode)
| Step |
Runs where |
Leaves your machine? |
| Parse files (Tree-sitter / fallback) |
Local process |
❌ No |
| Build code graph, entry/hub/flow insights |
Local |
❌ No |
| Mermaid diagrams & Mind Map |
Local webview |
❌ No |
| Git History (churn, owners, co-change) |
Local git CLI |
❌ No |
| Sidebar results & logs |
Local workspace state |
❌ No |
Full structural analysis and Git History run with no API key and no network LLM call.
🔐 Fully private LLM mode (recommended for sensitive repos)
- In the sidebar, choose Private codebase (or Command Palette →
NeverMIN: Use Private Codebase Mode).
- Install and run Ollama, then pick a model (
NeverMIN: Pilih Model Ollama).
- Cloud API keys are cleared automatically; switching to Gemini/OpenAI/etc. is blocked until you switch to Public mode.
With Ollama:
- Explain Code / insight narration talks to
127.0.0.1 (or your configured nevermin.ollamaBaseUrl) only.
- No cloud API key is required or kept for that session path.
- Prompts and code snippets are never sent to NeverMIN's authors or a hosted server — there is none.
⚠️ When code can leave your machine
Only if you choose a cloud provider (Gemini, OpenAI, Anthropic, OpenRouter, …) and run an LLM feature (Explain Code, narrative enrichment). The prompt — including selected code / insight summaries — then goes to that provider under their terms.
To stay leak-free: keep nevermin.provider = ollama, or skip LLM features entirely and use graph + Git History alone.
Secure setup checklist
1. nevermin.provider = ollama (auto when Private mode is chosen)
2. nevermin.ollamaBaseUrl = http://127.0.0.1:11434/v1
3. Do not Save API Key while on Ollama / Private mode
4. When finished: sidebar → Settings → Wipe this workspace data
Wipe this workspace data clears analysis, Git History, file checks, and LLM prompt cache from VS Code workspaceState for the current workspace — repo files are untouched. Choose Full reset to also re-pick Private/Public.
🚀 Quick Start
A. Install from the Marketplace (recommended)
|
VS Code
- Open Extensions (
Ctrl+Shift+X / Cmd+Shift+X)
- Search NeverMIN
- Click Install, reload if prompted
|
Cursor
Extensions → search NeverMIN → Install.
If it doesn't appear, use Install from VSIX.
|
Confirm it's installed: Extensions list shows NeverMIN enabled → NeverMIN icon appears in the Activity Bar → Command Palette lists NeverMIN: commands.
Use it on a project
- File → Open Folder… and open the codebase you want to understand.
- Click the NeverMIN icon in the Activity Bar.
- First run → choose Private codebase (local / Ollama only) or Public codebase (optional cloud LLM).
- In the sidebar: 2. Run (analyze whole repo) or 3. Select Files (analyze a selection).
- After analysis:
- 4. Structure → Folder → File → Function → click to open the Functions graph
- 5. Analysis Results → entry / hub / main flow / mind map
- 6. Git History → optional churn / owners / coupling (needs a git repo)
Optional LLM: Command Palette → pick a provider (prefer Ollama for private code).
B. Install from VSIX (local build, pre-release, or no Marketplace access)
Requirements: Node.js 18+, npm
git clone https://github.com/abdulwahidrukua/NeverMIN.git
cd NeverMIN
npm install
npm run package
Writes nevermin-<version>.vsix in the repo root.
UI: Extensions → … → Install from VSIX… → select the .vsix → reload if prompted.
CLI:
code --install-extension ./nevermin-0.0.2.vsix
# or, for Cursor:
cursor --install-extension ./nevermin-0.0.2.vsix
Update / uninstall
- Marketplace: Extensions → NeverMIN → Update / Uninstall
- VSIX: rebuild with
npm run package and reinstall (replaces previous version), or Uninstall from Extensions
C. Run from source (for hacking on NeverMIN itself)
git clone https://github.com/abdulwahidrukua/NeverMIN.git
cd NeverMIN
npm install
npm run compile
- Open the
NeverMIN folder in VS Code / Cursor.
- Press F5 (Run Extension) → an Extension Development Host window opens.
- In that window: File → Open Folder → select the project you want to analyze.
- Click the NeverMIN icon in the Activity Bar.
Development Host is for testing only — for daily use, prefer installing from the Marketplace.
🔄 Everyday Workflow
| 1 |
(Optional) Configure language & LLM — Command Palette (Ctrl+Shift+P / Cmd+Shift+P):
| Command |
Function |
NeverMIN: Pilih Bahasa / Choose Language |
UI in id or en |
NeverMIN: Pilih LLM Provider |
Gemini, OpenAI, Anthropic, OpenRouter, DeepSeek, Groq, Mistral, Together, xAI, Ollama (local) |
NeverMIN: Pilih Model Ollama |
Detect installed / loaded Ollama models |
NeverMIN: Simpan API Key |
Cloud providers only — stored in SecretStorage, cleared when you switch to Ollama |
Without a cloud key, graph analysis, diagrams, and structural Git History still work.
|
| 2 |
Analyze the repo — in the sidebar: check Status, optionally check off files in Select Files, then Run (whole repo or selection).
Same via Command Palette: NeverMIN: Analyze Repo Structure · NeverMIN: Select Files to Analyze
|
| 3 |
Analyze Git History (sidebar only) — answers what's alive vs. frozen, why the code looks like this, who knows the area, and hidden coupling.
- Ensure the folder is a git repository.
- Run → Analyze Git History, or open 6. Git History → Run Git History analysis.
- Results stay in the sidebar: LLM summary (if configured) · Alive/Frozen · Commits · Owners · Hidden coupling
Command: NeverMIN: Analisis Git History
|
| 4 |
Read the graph results — 5. Analysis Results shows graph statistics, Main Flow, entry points/hubs, and narration (if LLM enabled).
| Command |
Content |
NeverMIN: Open Main Flow (Mermaid) |
Flowchart of the main data flow |
NeverMIN: Open Insight Graph |
Architecture / module / functions panel |
NeverMIN: Open Learning Mind Map |
Mind map of learning order |
NeverMIN: Open Insights Summary |
Insights narration |
NeverMIN: Buka Ringkasan Git History |
Git History narration |
Or use 4. Structure: expand Folder → File → Function, click to open the focused Functions graph. Click a node / insight chip to jump to the related file.
|
| 5 |
Explain a code snippet — select code in the editor (or focus the active file), then Command Palette → NeverMIN: Jelaskan Kode Ini.
For a private repo, prefer Ollama so the snippet never hits a cloud API.
|
⚙️ Settings
Open Settings and search nevermin, or run NeverMIN: Open Settings.
Privacy mode (Private / Public) is not a VS Code setting — it's chosen in the NeverMIN sidebar, per workspace. Private mode locks the provider to Ollama and clears cloud API keys.
| Setting |
Default |
Description |
nevermin.language |
id |
UI + LLM language (id | en) |
nevermin.provider |
ollama |
LLM provider — Private mode forces ollama; Public allows cloud providers |
nevermin.model |
(empty) |
Model override; empty = provider default. For Ollama prefer NeverMIN: Pilih Model Ollama |
nevermin.temperature |
0.2 |
LLM sampling temperature |
nevermin.ollamaBaseUrl |
http://127.0.0.1:11434/v1 |
Ollama OpenAI-compatible base URL |
nevermin.maxAnalysisFiles |
500 |
Cap on files per analysis run |
API keys are not a VS Code setting — use NeverMIN: Simpan API Key (SecretStorage, per provider). Switching to Private / Ollama clears cloud keys automatically.
The sidebar Pengaturan panel mirrors the important actions: privacy mode, language, provider, API key / Ollama model, wipe workspace data.
📖 Commands Reference
Show all other commands
| Command |
Function |
NeverMIN: Refresh Sidebar |
Reload the sidebar tree |
NeverMIN: Buka Output Logs |
NeverMIN log channel |
NeverMIN: Bersihkan Logs |
Clear activity log |
NeverMIN: Centang Semua File / Kosongkan Centang File |
File selection for analysis |
NeverMIN: Pilih Model Ollama |
List & select local Ollama models |
🧩 Parsed Languages
Current scope: NeverMIN currently supports the JavaScript family (JS/JSX/TS/TSX) and Python only. Other languages are not yet parsed — support for more is on the roadmap.
|
Extension |
Parser |
 |
.js / .jsx |
Regex fallback |
 |
.ts |
Tree-sitter TypeScript |
 |
.tsx |
Tree-sitter TSX |
 |
.py |
Tree-sitter Python |
If the Tree-sitter WASM fails to load, symbol extraction still falls back and works.
🛠️ Development Scripts
| Script |
Function |
npm run compile |
Build src/ → dist/ |
npm run watch |
Auto-rebuild |
npm test |
Unit tests |
npm run copy-grammars |
Copy Tree-sitter WASM grammars |
npm run package |
Create .vsix |
npm install runs a postinstall step that copies grammars into media/grammars/.
