Survey of Software - Claude Skill#

Let Claude consult this research library directly in your conversations. The skill enables Claude to fetch surveys, synthesize recommendations, and run live research when topics aren’t covered yet.

Newsletter: New library research published as it’s ready. Monthly digest of what’s new — subscribe to stay current →

What It Does#

When you ask Claude about library or framework selection:

  • Fetches completed surveys from this research library
  • Synthesizes recommendations with decision frameworks and trade-offs
  • Runs rapid discovery research live when topics aren’t covered yet
  • Tailors recommendations to your specific requirements and constraints

Installation#

Option 1: Upload to Claude.ai#

  1. Download the skill:

  2. Upload to Claude:

    • Go to claude.ai → Settings → Skills
    • Click Upload skill
    • Upload the downloaded ZIP file

  3. Start using:

    • Create a new conversation in Claude.ai
    • The skill is now available for Claude to use automatically
Alternative: Package from source
git clone https://github.com/modelcitizendeveloper/survey-of-software.git
cd survey-of-software/skill
zip -r ../survey-of-software.zip .

Option 2: Use in Claude Code#

If you’re using the Claude Code CLI:

# Clone the repository
git clone https://github.com/modelcitizendeveloper/survey-of-software.git
cd survey-of-software

# Copy skill files to Claude Code skills directory
mkdir -p ~/.claude/skills/survey-of-software
cp skill/* ~/.claude/skills/survey-of-software/

The skill will be automatically available in all Claude Code sessions.

Example Queries#

Once installed, try asking Claude:

  • “What’s the best graph analysis library for Python?”
  • “Compare LLM orchestration frameworks for a RAG pipeline”
  • “What should I use for fuzzy search?”
  • “Which frontend framework should I choose for a new project?”
  • “I need a Chinese word segmentation library - what are my options?”
  • “What’s the current state of vector databases?”
  • “Compare state management libraries for React”

How It Works#

The skill uses two strategies depending on whether the topic has been researched:

For Covered Topics#

  1. Maps your question to the research taxonomy
  2. Fetches the relevant survey from this site
  3. Synthesizes a recommendation with:
    • Decision framework (when to use what)
    • Trade-off analysis (what you gain/lose with each choice)
    • Migration complexity considerations

For Uncovered Topics#

  1. Identifies the gap in the research library
  2. Offers to run S1 (Rapid Discovery) research live in your conversation
  3. Performs web research, evaluates options, presents findings following the survey format
  4. Invites contribution back to this open research project (CC BY 4.0)

Collaborative Four-Pass Model#

The skill implements the Four-Pass Survey (4PS) methodology:

  • S1: Rapid Discovery - Claude performs this automatically (70-80% confidence)
  • S2: Comprehensive Analysis - Available via deep research for benchmarks and detailed comparison
  • S3: Need-Driven Discovery - You provide your specific requirements, Claude evaluates fit
  • S4: Strategic Selection - You provide strategic context, Claude assesses long-term viability

Together, all four passes produce complete, tailored recommendations for your specific situation.

What’s Covered#

The skill currently covers the 1.xxx series (libraries):

  • Algorithms & Data Structures (1.001-1.049): Sorting, searching, graphs, strings, collections
  • Compression & Crypto (1.050-1.069): Compression, serialization, hashing, encryption
  • Machine Learning (1.070-1.079): Dimensionality reduction, forecasting, gradient boosting, deep learning
  • Image & Geometry (1.080-1.099): Image processing, QR codes, spatial algorithms, collision detection
  • Text Processing (1.100-1.109): PDF generation, document parsing, markdown, code formatting
  • Frontend & UI (1.110-1.119): Frameworks, state management, CSS, components, build tools, testing
  • Simulation & Business (1.120-1.139): Discrete event, Monte Carlo, financial simulation, CRM
  • Language Learning & CJK (1.140-1.169): Classical languages, spaced repetition, Chinese NLP, character databases
  • Translation (1.170-1.179): Machine translation, alignment, terminology
  • LLM & AI Stack (1.200-1.219): Orchestration, agents, vector databases, RAG, evaluation, fine-tuning
  • Calendar & Social (1.220-1.239): CalDAV, open social protocols, bot frameworks
  • Civic & Government (1.300-1.309): Public finance, government data, budget parsing, procurement

Future series: Standards (2.xxx) and Applications (3.xxx)

Source Code#

The skill is open source (CC BY 4.0):

Attribution#

When Claude presents findings using this skill, it will cite:

Source: Survey of Software by Model Citizen Developer. Four-Pass Survey (4PS) methodology. Licensed under CC BY 4.0.

Questions or Issues?#

  • Skill not working? Check that it’s uploaded correctly in Settings → Skills
  • Want to contribute research? The project is open source - PRs welcome
  • Found an error in a survey? Open an issue on GitHub
← License
Survey of Software - Research Library - © 2026 Ivan Schneider · Model Citizen Developer · Licensed under CC BY 4.0