cursor-docs（Cursor 文档）

Cursor Documentation Skill

CRITICAL: Path Doubling Prevention - MANDATORY

ABSOLUTE PROHIBITION: NEVER use cd with && in PowerShell when running scripts from this skill.

The Problem: If your current working directory is already inside the skill directory, using relative paths causes PowerShell to resolve paths relative to the current directory instead of the repository root, resulting in path doubling.

REQUIRED Solutions (choose one):

1. ALWAYS use absolute paths (recommended)
2. Use separate commands (never cd with &&)
3. Run from repository root with relative paths

NEVER DO THIS:

• Chain cd with &&: cd <relative-path> && python <script> causes path doubling
• Assume current directory
• Use relative paths when current dir is inside skill directory

CRITICAL: Large File Handling - MANDATORY SCRIPT USAGE

ABSOLUTE PROHIBITION: NEVER use read_file tool on the index.yaml file

The file exceeds context limits and will cause issues. You MUST use scripts.

REQUIRED: ALWAYS use manage_index.py scripts for ANY index.yaml access:

python scripts/management/manage_index.py count
python scripts/management/manage_index.py list
python scripts/management/manage_index.py get <doc_id>
python scripts/management/manage_index.py verify

All scripts automatically handle large files via index_manager.py.

Available Slash Commands

Use the consolidated docs-ops skill for common workflows:

• /cursor-ecosystem:docs-ops scrape - Scrape Cursor documentation from llms.txt sources
• /cursor-ecosystem:docs-ops refresh - Refresh the local index and metadata without scraping
• /cursor-ecosystem:docs-ops validate - Validate the index and references for consistency
• /cursor-ecosystem:docs-ops rebuild-index - Force rebuild the search index
• /cursor-ecosystem:docs-ops clear-cache - Clear the documentation search cache

Overview

This skill provides automation tooling for Cursor documentation management. It manages:

• Canonical storage (encapsulated in skill) - Single source of truth for official docs
• Subsection extraction - Token-optimized extracts (60-90% savings)
• Drift detection - Hash-based validation against upstream sources
• Sync workflows - Maintenance automation
• Documentation discovery - Keyword-based search and doc_id resolution
• Index management - Metadata, keywords, tags, aliases for resilient references

Core value: Prevents link rot, enables offline access, optimizes token costs, automates maintenance, and provides resilient doc_id-based references.

When to Use This Skill

This skill should be used when:

• Scraping documentation - Fetching docs from Cursor sources
• Finding documentation - Searching for docs by keywords, category, or natural language
• Resolving doc references - Converting doc_id to file paths
• Managing index metadata - Adding keywords, tags, aliases, updating metadata
• Rebuilding index - Regenerating index from filesystem (handles renames/moves)

Workflow Execution Pattern

CRITICAL: This section defines HOW to execute operations in this skill.

Delegation Strategy

Default approach: Delegate to Task agent

For ALL scraping, validation, and index operations, delegate execution to a general-purpose Task agent.

How to invoke:

Use the Task tool with:

• subagent_type: "general-purpose"
• description: Short 3-5 word description
• prompt: Full task description with execution instructions

Execution Pattern

Scripts run in FOREGROUND by default. Do NOT background them.

When Task agents execute scripts:

• Run directly: python plugins/cursor-ecosystem/skills/cursor-docs/scripts/core/scrape_docs.py
• Streaming logs: Scripts emit progress naturally via stdout
• Wait for completion: Scripts exit when done with exit code
• NEVER use run_in_background=true: Scripts are designed for foreground execution
• NEVER poll output: Streaming logs appear automatically, no BashOutput polling needed
• NEVER use background jobs: No &, no nohup, no background process management

Error and Warning Reporting

CRITICAL: Report ALL errors, warnings, and issues - never suppress or ignore them.

When executing scripts via Task agents:

• Report script errors: Exit codes, exceptions, error messages
• Report warnings: Deprecation warnings, import issues, configuration problems
• Report unexpected output: 404s, timeouts, validation failures
• Include context: What was being executed when the error occurred

Quick Start

Refresh Index End-to-End (No Scraping)

Use this when you want to rebuild and validate the local index/metadata without scraping:

python plugins/cursor-ecosystem/skills/cursor-docs/scripts/management/refresh_index.py

Scrape All Documentation

Use this when the user explicitly wants to hit the network and scrape docs:

# Scrape from configured llms.txt sources
python plugins/cursor-ecosystem/skills/cursor-docs/scripts/core/scrape_docs.py

# Refresh index after scraping
python plugins/cursor-ecosystem/skills/cursor-docs/scripts/management/refresh_index.py

Find Documentation

# Resolve doc_id to file path
python plugins/cursor-ecosystem/skills/cursor-docs/scripts/core/find_docs.py resolve <doc_id>

# Search by keywords (default: 25 results)
python plugins/cursor-ecosystem/skills/cursor-docs/scripts/core/find_docs.py search agent mcp

# Natural language search
python plugins/cursor-ecosystem/skills/cursor-docs/scripts/core/find_docs.py query "how to configure agent mode"

# List by category
python plugins/cursor-ecosystem/skills/cursor-docs/scripts/core/find_docs.py category core

# List by tag
python plugins/cursor-ecosystem/skills/cursor-docs/scripts/core/find_docs.py tag cli

Search Options:

Configuration System

The cursor-docs skill uses a unified configuration system.

Configuration Files:

• config/defaults.yaml - Central configuration file with all default values
• config/filtering.yaml - Content filtering rules
• config/tag_detection.yaml - Tag detection patterns

Environment Variable Overrides:

All configuration values can be overridden using environment variables: CURSOR_DOCS_<SECTION>_<KEY>

Dependencies

Required: pyyaml, requests, beautifulsoup4, markdownify, filelock Optional (recommended): yake (for keyword extraction)

Python Version: Python 3.11+ recommended

Core Capabilities

1. Scraping Documentation

Fetch documentation from configured sources using llms.txt format.

2. Extracting Subsections

Extract specific markdown sections for token-optimized responses.

3. Change Detection

Detect documentation drift via 404 checking and hash comparison.

4. Finding and Resolving Documentation

Discover and resolve documentation references using doc_id, keywords, or natural language queries.

5. Index Management and Maintenance

Maintain index metadata, keywords, tags, and rebuild index from filesystem.

Platform-Specific Requirements

Windows Users

MUST use PowerShell (recommended) or prefix Git Bash commands with MSYS_NO_PATHCONV=1

Git Bash on Windows converts Unix paths to Windows paths, breaking filter patterns.

Example:

MSYS_NO_PATHCONV=1 python scripts/core/scrape_docs.py --filter "/docs/"

Troubleshooting

Unicode Encoding Errors

Status: FIXED - Scripts auto-detect Windows and configure UTF-8 encoding.

404 Errors During Scraping

Status: EXPECTED - Some llms.txt entries may reference docs that don't exist yet. Scripts handle gracefully and continue.

Public API

The cursor-docs skill provides a clean public API for external tools:

from cursor_docs_api import (
    find_document,
    resolve_doc_id,
    get_docs_by_tag,
    get_docs_by_category,
    search_by_keywords,
    get_document_section,
    refresh_index
)

# Natural language search
docs = find_document("agent mode configuration")

# Resolve doc_id to metadata
doc = resolve_doc_id("cursor-getting-started")

# Get docs by tag
cli_docs = get_docs_by_tag("cli")

# Extract specific section
section = get_document_section("cursor-overview", "Installation")

Development Mode

When developing this plugin locally, you may want changes to go to your dev repo instead of the installed plugin location.

Enabling Dev Mode

PowerShell:

$env:CURSOR_DOCS_DEV_ROOT = "D:\repos\gh\melodic\claude-code-plugins"

Bash/Zsh:

export CURSOR_DOCS_DEV_ROOT="/path/to/claude-code-plugins"

Verifying Mode

When you run any major script (scrape, refresh, rebuild), a mode banner will display:

Dev mode:

[DEV MODE] Using local plugin: D:\repos\gh\melodic\claude-code-plugins

Prod mode:

[PROD MODE] Using installed skill directory

Disabling Dev Mode

PowerShell:

Remove-Item Env:CURSOR_DOCS_DEV_ROOT

Bash/Zsh:

unset CURSOR_DOCS_DEV_ROOT

Directory Structure

cursor-docs/
  SKILL.md                    # This file (public)
  cursor_docs_api.py          # Public API
  canonical/                  # Documentation storage (private)
    index.yaml                # Metadata index
  scripts/                    # Implementation (private)
    core/                     # Scraping, discovery
    management/               # Index management
    maintenance/              # Cleanup, drift detection
    utils/                    # Shared utilities
  config/                     # Configuration
    defaults.yaml             # Default settings
    filtering.yaml            # Content filtering
    tag_detection.yaml        # Tag patterns
  references/                 # Technical documentation (public)
    sources.json              # Documentation sources
  .cache/                     # Cache storage (inverted index)
  logs/                       # Log files

Source

Documentation is scraped from Cursor llms.txt at https://cursor.com/llms.txt.

Documentation Categories

Cursor documentation is organized into the following categories:

Tags

Common tags used for Cursor documentation:

• cli - Cursor CLI features
• ide - Cursor IDE features
• agent - Agent mode features
• tab - Tab completion features
• mcp - Model Context Protocol
• context - Context management
• enterprise - Enterprise features
• cloud-agent - Cloud Agent and API
• inline-edit - Inline editing features

Test Scenarios

Scenario 1: Keyword Search

Query: "Search for CLI documentation" Expected Behavior:

• Skill activates on keyword "documentation"
• Returns relevant docs from index Success Criteria: User receives matching documentation entries

Scenario 2: Natural Language Query

Query: "How do I configure the Cursor Agent mode?" Expected Behavior:

• Skill activates on "Cursor"
• Uses find_docs.py query command Success Criteria: Returns relevant documentation with configuration steps

Scenario 3: Doc ID Resolution

Query: "Resolve cursor-getting-started" Expected Behavior:

• Resolves doc_id to file path
• Returns document metadata Success Criteria: User receives full path and document content

Version History

• v1.0.0 (2025-12-16): Initial release - full skill structure adapted from codex-cli-docs

Last Updated

Date: 2025-12-16 Model: claude-opus-4-5-20251101

Status: Initial release - ready for scraping from https://cursor.com/llms.txt.