Loading

Sourceforge scores 73 out of 100 on the Agent Readiness Index

Sourceforge reaches level 2, Legible. Discoverability is the strongest area at 89 out of 100. Protocol access is the weakest at 36.

Sourceforge was scanned on November 14, 2026. Its score is 73 out of 100, level 2, Legible. An agent can read and cite your content. To reach level 3, Connectable: Offer a programmatic path: a live MCP server, a data sharing program with OAuth metadata, or a verified Poppy file.

Website: sourceforge.net

Score by area

AreaScoreWhat it asks
Discoverability89 out of 100, worth 22 pointsCan an agent find and read your pages?
Protocol access36 out of 100, worth 14 pointsCan an agent connect to you through MCP or an API?
Identity and delegation81 out of 100, worth 12 pointsCan a customer give an agent limited access without sharing a password?
Web experience60 out of 100, worth 20 pointsDoes your site work when an agent uses it?
Voice and IVRNot measured yetWhat happens when an agent calls your phone line?
Policy and governanceNot measured yetDo your terms allow agents that act for customers?
Marketing readiness48 out of 100, worth 18 pointsCan an agent see what you sell, and buy or book it?

What is holding Sourceforge back

  1. MCP server, 7.1 points lost. No MCP server, server card or registry entry was found for this organization. Fix: Publish a remote MCP server with OAuth. Start with read only tools such as account status, orders or location search, and publish a server card at /.well known/mcp/server card.
  2. Sign in page loads for a declared agent in a browser, 6.9 points lost. The sign in page did not load for an agent or a standard automated browser. Fix: Serve the sign in page to verified agent browsers, and use signed agent identity to separate them from attackers.
  3. An agent can start becoming a customer, 5.9 points lost. Book a demo: No link to book a demo on the home page. Fix: Make the call to action a plain link on the home page, keep the form behind it on a page an agent can load, label every field, and keep bot checks off the first step.
  4. Same behavior for a declared agent and a standard browser, 3.9 points lost. Both loads were blocked. Fix: Do not rank declared agents below anonymous browsers in bot management policy.

What already works

  • OAuth and OpenID metadata
  • PKCE and client registration
  • ChatGPT names what you sell

How Sourceforge compares

The average score for software is 62 out of 100 across 142 organizations. The average across all 2,353 scored organizations is 44.2 out of 100.

Similar organizations

More on this report