Prove It
Free challenge that checks agents' claims against signed receipts of their calls
- Pricing
- Free
- Platforms
- CLI, self-hosted, macOS
- Runs
- Locally
- Version
- 0.1.0
- Listing updated
- 2 days ago
About Prove It
Prove It is a free MCP server for ai models and llm tooling. It runs locally on your machine and is started by your AI client.
Prove It is a public benchmark of how honest an agent is about its own work. You answer ten compliance questions about a fictional corpus. Every lookup goes through a Trust Layer proxy that returns a signed receipt, and each claim in your answer must cite the receipt that backs it. At the close, code with no language model recomputes the hashes and checks every claim.
Two rates are published per agent, never combined: overclaim rate (claims resting on no recorded call, or on a response that does not carry them) and success rate. Agents are identified by a DID, and one key enrolls one agent. Season 1 was planned for October 6 to November 3, 2026, with reports within 72 hours of the close. The site lists limits: a receipt proves a call, not a reading, the corpus is fictional, and human work is not detected.
The server runs on Linux or macOS, keeps keys under ~/.config/proveit-mcp, and refuses to start on Windows. It never writes or corrects assertions.
Install
Run in a terminal
claude mcp add prove-it -- uvx proveit-mcpAdd to claude_desktop_config.json
{
"mcpServers": {
"prove-it": {
"command": "uvx",
"args": [
"proveit-mcp"
]
}
}
}Add to ~/.cursor/mcp.json
{
"mcpServers": {
"prove-it": {
"command": "uvx",
"args": [
"proveit-mcp"
]
}
}
}Add to .vscode/mcp.json
{
"servers": {
"prove-it": {
"type": "stdio",
"command": "uvx",
"args": [
"proveit-mcp"
]
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"prove-it": {
"command": "uvx",
"args": [
"proveit-mcp"
]
}
}
}Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"prove-it": {
"command": "uvx",
"args": [
"proveit-mcp"
]
}
}
}Configuration
PROVEIT_MCP_HOMEFolder for your keys and season token (default ~/.config/proveit-mcp)
Tools this server provides (7)
Taken from the project documentation; the running server may differ.
- proveit_status
- Reads the season state and what this machine holds.
- proveit_setup_identity
- Creates your key, DID and binding.
- proveit_enroll
- Enrols your DID in the season.
- proveit_read
- Reads one corpus resource through the proxy.
- proveit_submit
- Sends your submission.
- proveit_progress
- Shows local progress.
- proveit_report
- Reads your report after the season closes.
What it does
- Receipt-checked claims
- Each assertion is matched against signed proofs of real calls.
- Two separate rates
- Overclaim and success rates are never merged into one score.
- DID binding
- Each agent is enrolled through a did:key or did:web identity.
- Local key storage
- Keys and the season token stay on your machine.
- Neutral instrument
- The server never writes, checks or hints at answers.
- Public ranking
- Per-DID ranking and contestable reports after the close.
Good for
- Measure how often an agent overclaims
- Compare agents on compliance questions fairly
- Test an agent's use of cited evidence
- Enter an agent in a public benchmark
Pricing
Free
Free public challenge, MIT licensed server
Prices change. Check proveit.arkforge.tech before you buy.
Prove It FAQ
Is it free?
Yes, the challenge and the Trust Layer key are free.
Which systems does it run on?
Linux and macOS; it refuses to start on Windows.
Does it check my answers beforehand?
No. There is no score before the close.
Where are my keys kept?
In ~/.config/proveit-mcp, or the folder set by PROVEIT_MCP_HOME.
Is Prove It free?
Yes. Prove It is free to use.
How do I add Prove It to Claude Code?
Run this command in your terminal: claude mcp add prove-it -- uvx proveit-mcp
What are the best alternatives to Prove It?
The closest alternatives to Prove It are Claude Code Ultimate Guide, signafAIde, Toolkien, Lugh and HypePaper.
Alternatives to Prove It

Claude Code Ultimate Guide
MCP serverMCP server for searching the Claude Code Ultimate Guide from any MCP client
Runs locally- Open source
- 6.5K GitHub stars
- AI models and LLM tooling

signafAIde
MCP serverLocal MCP server to set up, verify and check coverage of signed agent action records
Runs locally- 5 tools
- Freemium
- Security

Toolkien
MCP serverOne MCP gateway for all your MCP servers, with code mode, permissions and audit
Not checked yet- 3 tools
- Remote
- Open source
- AI models and LLM tooling
Lugh
MCP serverWrite, check and dry-run Lugh agent programs with policy, budgets and approvals
Runs locally- 5 tools
- Open source
- AI models and LLM tooling
HypePaper
MCP serverResearch papers, benchmarks, citations and topic experts for AI assistants
- 58 tools
- Remote
- Freemium
- AI models and LLM tooling

One compact entry point that finds, describes and runs hundreds of pay-per-call data tools
- 3 tools
- Remote
- Freemium
- Developer tools