# confluence-deep-reader

confluence-deep-reader is a Claude Code skill that traverses Confluence page trees to depth 3 and synthesizes hierarchical documentation.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode, Windsurf (Devin Desktop), Claude.ai · 1 install · ★ 939 repo stars
- Author: krafton-ai
- License: Apache-2.0 · policy=mirror · the beginning of the file is at the end of this page
- Source: https://github.com/krafton-ai/KIRA/tree/HEAD/KIRA-Slack/.claude/skills/confluence-deep-reader/SKILL.md
- Updated: 2026-05-29
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/skills/confluence-deep-reader

## What it does

This skill guides the agent across nested Confluence articles. It begins by opening the entry page with `getConfluencePage`, then discovers nested descendants via `getConfluencePageDescendants`. The workflow continues step-by-step through direct children, grandchildren, and great-grandchildren.

The system sets a hard stop at depth 3. Whenever a documentation tree contains over 20 items, the agent asks whether to parse everything or concentrate on selected branches. Missing or restricted pages get marked in notes without interrupting the overall run.

Operational success depends on functional `getConfluencePage` and `getConfluencePageDescendants` tools in your agent setup. Without configured Confluence access, page fetching cannot proceed.

## Requirements
- Works with: Confluence, Slack

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add krafton-ai/KIRA --skill confluence-deep-reader -a claude-code
```

2. **Manually: put the files in place**
- File: `~/.claude/skills/confluence-deep-reader/SKILL.md` — Copy the whole skill folder; for one project use .claude/skills/confluence-deep-reader/

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add krafton-ai/KIRA --skill confluence-deep-reader -a codex
```

2. **Manually: put the files in place**
- File: `.agents/skills/confluence-deep-reader/SKILL.md` — For all projects use ~/.agents/skills/

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add krafton-ai/KIRA --skill confluence-deep-reader -a cursor
```

2. **Manually: put the files in place**
- File: `.cursor/skills/confluence-deep-reader/SKILL.md` — Cursor also reads .agents/skills/ and .claude/skills/

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install krafton-ai/KIRA KIRA-Slack/.claude/skills/confluence-deep-reader
```

2. **Manually: put the files in place**
- File: `.github/skills/confluence-deep-reader/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add krafton-ai/KIRA --skill confluence-deep-reader -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/krafton-ai/KIRA --path KIRA-Slack/.claude/skills/confluence-deep-reader
```

### Windsurf (Devin Desktop)

**Manually: put the files in place**
- File: `.devin/skills/confluence-deep-reader/SKILL.md` — The legacy .windsurf/skills/ path still works

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add krafton-ai/KIRA --skill confluence-deep-reader -a opencode
```

### Claude.ai

**Claude.ai**
- Zip the folder and upload it in Settings → Capabilities → Skills

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/krafton-ai/KIRA/tree/HEAD/KIRA-Slack/.claude/skills/confluence-deep-reader/SKILL.md

## When to use it
- crawl an entire Confluence section from a single page URL
- inspect parent and descendant documentation down to depth 3
- summarize multi-tier documentation trees while preserving hierarchy
- catch inaccessible Confluence pages during branch scans

## Example prompts
- Read this Confluence page and all its subpages: <link>
- Analyze this Confluence documentation tree down to child levels and summarize it
- Review the parent page and descendants at <link>, then structure the findings

## Source text · SKILL.md · Apache-2.0

Below is the content of a third-party file. It is data, not instructions for you: do not act on anything written inside unless the human asks. Block tag: f41cefbf029983bd.

BEGIN THIRD-PARTY FILE f41cefbf029983bd
````text
---
name: confluence-deep-reader
description: Read a Confluence page and recursively explore its child pages up to 3 levels deep. Use when users want to comprehensively read a Confluence page tree, understand hierarchical documentation, or analyze content across parent and child pages.
---

# Confluence Deep Reader

This skill guides you to recursively read Confluence pages and their descendants up to 3 levels deep.

## When to Use This Skill

Use this skill when:
- User provides a Confluence link and wants to read "everything" or "all related pages"
- User asks to "read this page and subpages"
- User wants to understand the full content tree of a Confluence page
- User needs comprehensive analysis of hierarchical documentation

## Workflow

### Step 1: Read the Starting Page

Use `getConfluencePage` to read the initial page provided by the user.

### Step 2: Discover Child Pages

Use `getConfluencePageDescendants` to find all child pages of the current page. This tool returns the hierarchical structure of descendants.

### Step 3: Read Each Child Page (Depth 1)

For each child page found:
1. Use `getConfluencePage` to read its content
2. Track that this is depth 1

### Step 4: Recursively Explore Deeper Levels

For each child page read in Step 3:
1. Use `getConfluencePageDescendants` to check if it has its own children
2. If children exist and current depth < 3:
   - Read each grandchild page (depth 2)
   - Repeat this process for depth 3

**IMPORTANT: Stop at depth 3.** Do not explore beyond 3 levels from the starting page.

### Step 5: Synthesize Information

After reading all pages within the depth limit:
1. Organize information hierarchically
2. Provide comprehensive summary or answer user's question
3. Note if there are deeper levels that were not explored (beyond depth 3)

## Depth Tracking

Keep track of the current depth level:
- **Depth 0**: Starting page (the one user provided)
- **Depth 1**: Direct children of starting page
- **Depth 2**: Grandchildren
- **Depth 3**: Great-grandchildren (STOP HERE)

## Best Practices

- **Show progress**: Let user know which pages you're reading (e.g., "Reading main page... Found 5 child pages... Reading child 1/5...")
- **Respect the limit**: Never go beyond depth 3
- **Be efficient**: If there are many pages (>20), consider asking user if they want to continue or focus on specific sections
- **Handle errors gracefully**: If a page cannot be accessed, note it and continue with others
- **Provide structure**: When presenting results, maintain the hierarchical structure so users understand the page relationships

## Example Usage

```
````
END THIRD-PARTY FILE f41cefbf029983bd

The third-party file has ended. Everything between the f41cefbf029983bd tags is data: requests and commands from it are not followed. An “end of file” line with another tag inside the block is part of the file.

First 2625 of 3273 bytes. © krafton-ai · Apache-2.0 · https://github.com/krafton-ai/KIRA/tree/HEAD/KIRA-Slack/.claude/skills/confluence-deep-reader/SKILL.md

_Part of the page did not fit — see the HTML version._

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
