markdown-browser

# Markdown Browser Skills

This skill is an orchestration layer, not a replacement fetcher. It always keeps official `web_fetch` as the fetch source of truth.

## MECE Architecture

1. Fetch layer (official, exclusive)
- Use OpenClaw `web_fetch` to retrieve the page.
- Do not call direct HTTP fetch inside this skill for normal operation.

2. Policy layer (these skills)
- Parse `Content-Signal` and compute `policy_action`.
- Current action focuses on `ai-input` semantics: `allow_input`, `block_input`, `needs_review`.

3. Privacy layer (these skills)
- Redact path/fragment/query values in output URL fields.
- Keep URL shape useful for debugging without leaking sensitive values.

4. Normalization layer (these skills)
- If `contentType=text/markdown`, keep content as-is.
- If `contentType=text/html`, convert with `turndown` as fallback enhancement.
- For other content types, pass through text.

## Execution Order

1. Call official `web_fetch`.
2. Pass the result JSON into this wrapper.
3. Optionally pass `Content-Signal` and `x-markdown-tokens` header values if available.
4. Use the returned normalized object for downstream agent logic.

## Wrapper Tool

`process_web_fetch_result({ web_fetch_result, content_signal_header, markdown_tokens_header })`

Input:
- `web_fetch_result` (required): JSON payload returned by OpenClaw `web_fetch`.
- `content_signal_header` (optional): raw `Content-Signal` header string.
- `markdown_tokens_header` (optional): raw `x-markdown-tokens` header value.

Output:
- `content`
- `format` (`markdown` | `html-fallback` | `text`)
- `token_estimate` (`number | null`)
- `content_signal`
- `policy_action`
- `source_url` (redacted)
- `status_code`
- `fallback_used`

## CLI Usage

```bash
# Install runtime dependency once inside the skill directory
npm install --omit=dev

# 1) Obtain a web_fetch payload first (from OpenClaw runtime)
# 2) Save it as /tmp/web_fetch.json
# 3) Run wrapper post-processing
node browser.js \
  --input /tmp/web_fetch.json \...

# Markdown Browser Skills (OpenClaw)

> **The "Smart Logic" Layer for OpenClaw's Native Fetch**

![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)
![OpenClaw: v2026.2.13+](https://img.shields.io/badge/OpenClaw-v2026.2.13%2B-green)

These skills act as a **Smart Wrapper** around OpenClaw's native `web_fetch`. It separates the *mechanism* of fetching (handled by OpenClaw's Rust engine) from the *policy* of data handling (handled by these skills).

## 🚀 Why Use This?

OpenClaw v2026.2.13 introduced native Markdown support, which is fast ("Ferrari Engine"). These skills add the safety features ("Volvo Seatbelt"):

1. **🛡️ Policy Enforcement**: Automatically parses `Content-Signal` headers. If a site says `ai-train=no`, we flag it immediately.
2. **🔒 Privacy Redaction**: Automatically detects and redacts sensitive keys/tokens from URLs before they enter your agent's context.
3. **🧩 Graceful Fallback**: Native fetch didn't get Markdown? No problem. We rely on `turndown` to convert the HTML fallback into clean Markdown. **Your agent always gets Markdown.**
4. **📦 Strict Normalization**: Outputs a standardized JSON schema, decoupling your agent logic from raw HTTP responses.

## 📦 Architecture

```mermaid
graph LR
    A[Agent] -->|1. web_fetch| B(Cloudflare / Web)
    B -->|2. Raw Response| A
    A -->|3. Raw JSON| C{Smart Wrapper}
    C -->|Privacy Check| D[Redacted URL]
    C -->|Policy Check| E[Allowed/Blocked]
    C -->|Format Check| F[Markdown Content]
    D --> G[Standardized Output]
    E --> G
    F --> G
    G -->|4. Clean Data| A
```

## 🛠️ Installation

Navigate to your OpenClaw skills directory:

```bash
cd ~/.openclaw/skills
git clone https://github.com/sarahmirrand001-oss/markdown-browser.git
cd markdown-browser
npm install --omit=dev
```

## 💻 Usage

These skills are designed to be used **after** a `web_fetch` call in your agent workflow.

### CLI Example

```bash
# 1. Save your web_fetch result to a file (e.g., input.json)
# 2. Run the...