A local-first, lightweight Markdown workflow tool for AI-assisted development.
Turn plain .md files into interactive pages with checkboxes, progress bars, Mermaid diagrams, editable text blocks, buttons, and write-back updates.
start_md.bator
bash start_md.shThen open http://127.0.0.1:8000
Reference http://127.0.0.1:8000/development_progress.md
To use a different port, pass --port <port>
To use a different folder, pass --cd <folder>
MD Activator serves Markdown files from a local folder and renders them as interactive browser pages.
It supports:
- Markdown viewing
- Mermaid diagram rendering
- image and link rendering
- task checkboxes with write-back updates
- progress bars generated from task lists
- editable text blocks with save-back behavior
- button-style checklist actions
- AI-assisted development notes
- human-in-the-loop task workflows
- requirement analysis documents
- implementation progress tracking
- lightweight local project dashboards
- Markdown-based issue execution plans
MD Activator is designed for local use. It reads and updates Markdown files from the selected content folder.
Do not expose the server directly to the public internet unless you have added proper authentication, authorization, and network-level protection.
- The server reads markdown files from the folder where you start it unless a content root override is provided.
- Default file is `README.md`.
- Relative links ending in `.md` are opened inside the app.
- Table, URL, Image, Mermaid will be rendered as html page components. File name will be rendered as url and downloadable.
- Double click a text box enable editing the text. After editing, mouse click outside of the editor will auto save
the changes back to the original .md file.
- Task markers like `[ ]` and `[x]` at the start of a line are rendered as checkboxes and checkable. The check/uncheck
value will be updated to the original .md file.
- `[ ]` and `[x]` followed with "[<text>]" be rendered as button with check indicator. Click the button will switch the
check icon display, and the change will be updated to the original .md file.
- `[ ]` and `[x]` with "progress" in the above line will be rendered as step progress bar and read only.
Pass --cd <folder> to serve markdown from another folder:
start_md.bat run python -m app.main --cd /path/to/notes --reloadWhen starting Uvicorn directly, set MD_VIEWER_ROOT before starting the server:
MD_VIEWER_ROOT=/path/to/notes uvicorn app.main:app --reloadContent root precedence is --cd, then MD_VIEWER_ROOT, then the process working directory.
From PowerShell or CMD:
start_md.batThe launcher uses uv sync --quiet to create/update .venv from pyproject.toml,
then starts the app with uv run --no-sync.
Optional: serve a specific folder:
start_md.bat C:\path\to\notesOptional: start on a different port:
start_md.bat -p 8124 C:\path\to\notes| Library | Good with Quasar? | Main use |
|---|---|---|
| Bootstrap | Usually no | Overlapping UI framework |
| Mermaid.js | Yes | Diagrams |
| PrismJS | Yes | Code highlighting |
Test content 2
[] Option 1
[x] Option 2
[] Option 3
[x] Option 4
progress
[X] Step 1 Activate the code and do something
[x] Step 2 Dectivate the code and do something
[] Step 3
[] Step 4
[] Step 5
[] Step 6
[] [confirm]
- Primary item
- Secondary item
- Nested item
1. First operation
2. Second operation
[GitHub](https://github.com)
img/website_bird_protection.png
```mermaid
flowchart TD
A[User enters search query] --> B[Search Frontend]
B --> C[Query Understanding]
C --> C1[Spell correction]
C --> C2[Intent detection]
C --> C3[Entity recognition]
C --> C4[Language and location context]
C --> D[Retrieve Candidate Pages]
subgraph Indexing Pipeline
E[Web Crawlers / Bot] --> F[Fetch web pages]
F --> G[Parse HTML, links, media, metadata]
G --> H[Render JavaScript when needed]
H --> I[Extract content and signals]
I --> J[Build Search Index]
end
```