Feature Support Matrix & Platform Comparison

Complete comparison of Skill Seekers features across all supported platforms.

Platform Overview

Skill Seekers supports 4 target platforms with different formats, APIs, and capabilities:

Platform	Format	API	Enhancement	Status
Claude AI	ZIP, YAML frontmatter	Skills API (manual)	✅ LOCAL + API	✅ Production
Google Gemini	tar.gz, plain markdown	Files API	✅ API only	✅ Production
OpenAI ChatGPT	ZIP, assistant instructions	Assistants API	✅ API only	✅ Production
Generic Markdown	ZIP, universal markdown	N/A	❌ Not applicable	✅ Production

Platform Capabilities

File Formats

Platform	Package Format	SKILL.md Format	References
Claude AI	`.zip`	Markdown + YAML frontmatter	`references/` directory
Google Gemini	`.tar.gz`	Plain markdown (no frontmatter)	`references/` directory
OpenAI ChatGPT	`.zip`	Plain text (assistant instructions)	`vector_store_files/`
Generic Markdown	`.zip`	Universal markdown	`references/` directory

Command:

skill-seekers package output/skill/ --target [claude|gemini|openai|markdown]

Enhancement Support

Platform	API Mode	LOCAL Mode	Cost	Speed
Claude AI	✅ Yes (Anthropic API)	✅ Yes (Claude Code Max)	$0.15-0.30 / FREE	Fast / Moderate
Google Gemini	✅ Yes (Gemini API)	❌ No	$0.01-0.05	Fast
OpenAI ChatGPT	✅ Yes (OpenAI API)	❌ No	$0.15-0.30	Fast
Generic Markdown	❌ N/A	❌ N/A	N/A	N/A

Commands:

# Claude (API mode)
export ANTHROPIC_API_KEY=sk-ant-...
skill-seekers enhance output/skill/ --enhance

# Claude (LOCAL mode - FREE!)
skill-seekers enhance output/skill/ --enhance-local

# Gemini (API mode)
export GOOGLE_API_KEY=AIza...
skill-seekers enhance output/skill/ --target gemini

# OpenAI (API mode)
export OPENAI_API_KEY=sk-proj-...
skill-seekers enhance output/skill/ --target openai

Upload Methods

Platform	Manual Upload	Programmatic Upload	MCP Tool
Claude AI	✅ claude.ai/skills	🚧 Coming soon	🚧 Coming soon
Google Gemini	✅ aistudio.google.com	✅ Python API	✅ Yes
OpenAI ChatGPT	✅ platform.openai.com	✅ Python API	✅ Yes
Generic Markdown	❌ N/A	❌ N/A	❌ N/A

Commands:

# Gemini (programmatic)
skill-seekers upload skill-gemini.tar.gz --target gemini --api-key AIza...

# OpenAI (programmatic)
skill-seekers upload skill-openai.zip --target openai --api-key sk-proj-...

# Claude (manual for now)
# Upload skill.zip at claude.ai/skills

Skill Modes

Single-Source Skills

All platforms support:

Mode	Command	Output	Platforms
Documentation	`skill-seekers scrape`	Single skill from docs	All
GitHub	`skill-seekers github`	Single skill from repo	All
PDF	`skill-seekers pdf`	Single skill from PDF	All
Codebase (C3.x)	`skill-seekers-codebase`	Single skill from code	All

Multi-Source Skills

Platform-specific support:

Mode	Command	Output	Claude	Gemini	OpenAI	Markdown
Unified	`skill-seekers unified`	Merged skill	✅ Yes	✅ Yes	✅ Yes	✅ Yes
Router (3-Stream)	`skill-seekers router`	Router + sub-skills	✅ Yes	⚠️ Limited	⚠️ Limited	❌ No

Notes:

Unified mode: All platforms merge multiple sources into one skill
Router mode: Claude AI has native sub-skill support; other platforms require manual sub-skill management

CLI Command Support

Scraping Commands

Command	Description	All Platforms
`skill-seekers scrape`	Documentation scraping	✅ Yes
`skill-seekers github`	GitHub repository	✅ Yes
`skill-seekers pdf`	PDF extraction	✅ Yes
`skill-seekers unified`	Multi-source scraping	✅ Yes
`skill-seekers-codebase`	Codebase analysis (C3.x)	✅ Yes

Processing Commands

Command	Description	Claude	Gemini	OpenAI	Markdown
`skill-seekers enhance`	AI enhancement	✅ API + LOCAL	✅ API only	✅ API only	❌ N/A
`skill-seekers package`	Package skill	✅ Yes	✅ Yes	✅ Yes	✅ Yes
`skill-seekers upload`	Upload to platform	🚧 Manual	✅ Yes	✅ Yes	❌ N/A
`skill-seekers validate`	Validate structure	✅ Yes	✅ Yes	✅ Yes	✅ Yes
`skill-seekers router`	Generate router	✅ Yes	⚠️ Limited	⚠️ Limited	❌ No

Legend:

✅ Full support
⚠️ Partial support
🚧 Coming soon
❌ Not applicable

C3.x Codebase Analysis

Feature Support

All platforms support C3.x codebase analysis features:

Feature	Command Flag	Output	All Platforms
Pattern Detection	`--extract-patterns`	Design patterns	✅ Yes
Test Extraction	`--extract-test-examples`	5 test categories	✅ Yes
How-To Guides	`--build-how-to-guides`	Tutorial generation	✅ Yes
Config Patterns	`--extract-config-patterns`	Config analysis	✅ Yes
Architecture	`--generate-architecture`	System overview	✅ Yes
AI Enhancement	`—ai-mode [auto	api	local

Example:

# Works for all platforms
skill-seekers-codebase path/to/repo/ \
  --output output/codebase/ \
  --extract-patterns \
  --extract-test-examples \
  --build-how-to-guides \
  --ai-mode auto

AI Enhancement Modes

Mode	Claude	Gemini	OpenAI	Markdown
AUTO (detect best)	✅ LOCAL if available, else API	✅ API	✅ API	❌ N/A
API	✅ Anthropic API	✅ Gemini API	✅ OpenAI API	❌ N/A
LOCAL	✅ Claude Code Max (FREE!)	❌ No	❌ No	❌ N/A
NONE	✅ Basic output	✅ Basic output	✅ Basic output	✅ Basic output

MCP Tool Support

Available MCP Tools

18 MCP tools for Claude Desktop integration:

Tool Category	Tool Name	Description
Scraping	`scrape_docs`	Documentation scraping
	`scrape_github`	GitHub repository
	`scrape_pdf`	PDF extraction
	`unified_scrape`	Multi-source scraping
Skill Management	`package_skill`	Package for platform
	`upload_skill`	Upload to Claude/Gemini/OpenAI
	`enhance_skill`	AI enhancement
	`validate_skill`	Validate structure
C3.x Analysis	`analyze_codebase`	Full codebase analysis
	`extract_patterns`	Pattern detection
	`extract_test_examples`	Test extraction
	`build_how_to_guides`	Tutorial generation
Git Sources	`add_git_source`	Add git config source
	`list_git_sources`	List sources
	`remove_git_source`	Remove source
	`fetch_git_sources`	Fetch updates
Utilities	`list_presets`	Show presets
	`get_preset`	Get preset config

Setup:

pip install skill-seekers[mcp]

# Add to Claude Desktop config
# ~/.config/Claude/claude_desktop_config.json (Linux)
# ~/Library/Application Support/Claude/claude_desktop_config.json (macOS)
# %APPDATA%/Claude/claude_desktop_config.json (Windows)

See also: MCP Setup Guide

Workflow Coverage

Workflow: Documentation → Skill

Step	Command	Claude	Gemini	OpenAI	Markdown
1. Scrape	`skill-seekers scrape`	✅	✅	✅	✅
2. Enhance	`skill-seekers enhance`	✅ API + LOCAL	✅ API	✅ API	❌
3. Package	`skill-seekers package --target X`	✅ ZIP	✅ tar.gz	✅ ZIP	✅ ZIP
4. Upload	`skill-seekers upload --target X`	🚧 Manual	✅ API	✅ API	❌

Workflow: GitHub → Skill

Step	Command	Claude	Gemini	OpenAI	Markdown
1. Scrape	`skill-seekers github`	✅	✅	✅	✅
2. Package	`skill-seekers package --target X`	✅ ZIP	✅ tar.gz	✅ ZIP	✅ ZIP
3. Upload	`skill-seekers upload --target X`	🚧 Manual	✅ API	✅ API	❌

Workflow: Codebase (C3.x) → Skill

Step	Command	Claude	Gemini	OpenAI	Markdown
1. Analyze	`skill-seekers-codebase`	✅	✅	✅	✅
2. Extract Patterns	`--extract-patterns`	✅	✅	✅	✅
3. Build Guides	`--build-how-to-guides`	✅	✅	✅	✅
4. AI Enhance	`--ai-mode auto`	✅ LOCAL + API	✅ API	✅ API	❌
5. Package	`skill-seekers package --target X`	✅ ZIP	✅ tar.gz	✅ ZIP	✅ ZIP
6. Upload	`skill-seekers upload --target X`	🚧 Manual	✅ API	✅ API	❌

Workflow: Multi-Source (Unified) → Skill

Step	Command	Claude	Gemini	OpenAI	Markdown
1. Unified Scrape	`skill-seekers unified`	✅	✅	✅	✅
2. Conflict Detection	Automatic	✅	✅	✅	✅
3. Merge	`--merge-mode rule-based`	✅	✅	✅	✅
4. Enhance	`skill-seekers enhance`	✅ API + LOCAL	✅ API	✅ API	❌
5. Package	`skill-seekers package --target X`	✅ ZIP	✅ tar.gz	✅ ZIP	✅ ZIP
6. Upload	`skill-seekers upload --target X`	🚧 Manual	✅ API	✅ API	❌

Workflow: Three-Stream Router → Skill

Step	Command	Claude	Gemini	OpenAI	Markdown
1. Code Stream	`skill-seekers-codebase`	✅	✅	✅	✅
2. Docs Stream	`skill-seekers scrape`	✅	✅	✅	✅
3. GitHub Stream	`skill-seekers github`	✅	✅	✅	✅
4. Generate Router	`skill-seekers router`	✅ Full support	⚠️ Single skill	⚠️ Single skill	❌
5. Package All	`--include-subskills`	✅ ZIP	✅ tar.gz	✅ ZIP	❌
6. Upload	`skill-seekers upload --target X`	🚧 Manual	✅ API	✅ API	❌

Note: Gemini and OpenAI don’t have native sub-skill routing, so the router flattens into a single comprehensive skill.

Platform-Specific Features

Claude AI Exclusive

✅ LOCAL Enhancement Mode - FREE with Claude Code Max ✅ Sub-Skill Routing - Native hierarchical skill support ✅ Custom Instructions - Behavior customization in YAML frontmatter ✅ MCP Integration - 18 tools for Claude Desktop

Google Gemini Exclusive

✅ Grounding Support - Automatic source attribution ✅ Long Context Window - Up to 1M tokens ✅ Low Cost Enhancement - $0.01-0.05 per skill (Gemini 2.0 Flash)

OpenAI ChatGPT Exclusive

✅ Vector Store + file_search - Semantic search built-in ✅ Function Calling - Extend with custom tools ✅ Streaming Responses - Real-time answer generation

Generic Markdown

✅ Universal Compatibility - Works with any LLM ✅ No Platform Lock-In - Pure markdown format ✅ Version Control Friendly - Easy to diff and track changes

Preset Support

24 built-in presets work across all platforms:

Category	Presets	All Platforms
Game Engines	Godot, Unity, Unreal	✅ Yes
Web Frameworks	React, Vue, Angular, Svelte, Next.js	✅ Yes
Backend	Django, FastAPI, Flask, Express, NestJS	✅ Yes
Languages	Python, TypeScript, Rust, Go	✅ Yes
Tools	Tailwind, Kubernetes, Docker, PostgreSQL	✅ Yes

Usage:

# List all presets
skill-seekers list-presets

# Use preset (works for all platforms)
skill-seekers scrape --preset react
skill-seekers package output/react/ --target [claude|gemini|openai|markdown]

Cost Comparison

Enhancement Costs

Platform	Model	Input Cost	Output Cost	Typical Skill	Total Cost
Claude AI (API)	Claude Sonnet	$3/1M	$15/1M	50K in, 10K out	$0.30
Claude AI (LOCAL)	Claude Code Max	FREE	FREE	Any size	$0.00 🎉
Google Gemini	Gemini 2.0 Flash	$0.075/1M	$0.30/1M	50K in, 10K out	$0.01
OpenAI ChatGPT	GPT-4o	$2.50/1M	$10.00/1M	50K in, 10K out	$0.23

Storage/Upload Costs

Platform	Storage Cost	Upload Cost	Notes
Claude AI	FREE	FREE	Manual upload (for now)
Google Gemini	$0.10/GB/day	FREE	Files API
OpenAI ChatGPT	$0.10/GB/day	FREE	Vector Store
Generic Markdown	N/A	N/A	No upload

Version Compatibility

Feature	Version Added	Claude	Gemini	OpenAI	Markdown
Documentation Scraping	v1.0.0	✅	✅	✅	✅
GitHub Scraping	v1.5.0	✅	✅	✅	✅
PDF Extraction	v2.0.0	✅	✅	✅	✅
Multi-LLM Support	v2.5.0	✅	✅	✅	✅
C3.x Codebase Analysis	v2.6.0	✅	✅	✅	✅
Three-Stream Architecture	v2.6.0	✅	⚠️	⚠️	❌
AI Enhancement (API)	v2.5.0	✅	✅	✅	❌
AI Enhancement (LOCAL)	v2.6.0	✅	❌	❌	❌
Unified Scraping	v2.0.0	✅	✅	✅	✅
MCP Integration	v2.5.0	✅	⚠️	⚠️	❌

Platform Selection Guide

Choose Claude AI if:

✅ You have Claude Code Max (FREE LOCAL enhancement!)
✅ You need sub-skill routing (router pattern)
✅ You want MCP integration with Claude Desktop
✅ You prefer YAML frontmatter structure

Choose Google Gemini if:

✅ You want lowest enhancement cost ($0.01-0.05)
✅ You need very long context window (1M tokens)
✅ You want automatic grounding/citations
✅ You prefer plain markdown format

Choose OpenAI ChatGPT if:

✅ You need Vector Store + semantic search
✅ You want function calling capabilities
✅ You need streaming responses
✅ You’re building custom GPTs

Choose Generic Markdown if:

✅ You want platform independence
✅ You need version control friendly format
✅ You’re using custom/local LLMs
✅ You want maximum portability

Next Steps

Multi-LLM Support Guide - Detailed platform comparison
Claude Integration - Claude-specific features
Gemini Integration - Gemini setup and usage
OpenAI Integration - OpenAI setup and usage

Status: ✅ Complete (v2.6.0)

Found an issue or have suggestions? Open an issue