outputproof-server

OutputProof Server

Source-available dashboard server and team API for OutputProof.

This package contains the monetizable server/dashboard surface described in the OutputProof product requirements:

• local and team dashboard API
• verification history views
• agent reliability scorecards
• team reliability leaderboard across agents, developers, and task types
• failure pattern analysis foundations
• SQLite-backed local persistence with no external database required
• future policy engine and cloud sync hooks

The core SDK, assertion engine, LLM-as-judge module, CLI, and MCP integration remain Apache 2.0 in outputproof.

Where Dashboard Data Comes From

The dashboard shows verification records stored by outputproof-server. Records are created through the server API:

POST /api/verifications

In normal local use, the CLI sends those records when OUTPUTPROOF_SERVER_URL points at the running server. In production/team use, CI jobs, agent runners, MCP integrations, or application code send the same VerificationResult JSON to the API.

The v1.1 GitHub Actions gate includes developer_id, task_type, changed-file, and GitHub run metadata in each verification record. The server uses that metadata to populate the team reliability leaderboard:

GET /api/leaderboard?dimension=agent
GET /api/leaderboard?dimension=developer
GET /api/leaderboard?dimension=task_type

The server does not scan your workspace and does not automatically import the SDK local history file at ~/.outputproof/verifications.jsonl; that local file is the Apache SDK audit log. The BSL server dashboard has its own SQLite database, defaulting to ~/.outputproof/outputproof-server.db.

Usage

Install the Apache SDK first, then install this server package from the sibling directory:

# From C:\workspace\ai-agent-output-verification
cd outputproof-sdk
python -m pip install -e .

cd ..\outputproof-server
python -m pip install -e .

Run the dashboard server:

python -m outputproof_server.cli --port 8080
# Then open http://127.0.0.1:8080

# Optional: choose a specific SQLite file
python -m outputproof_server.cli --database ~/.outputproof/outputproof-server.db

If PowerShell says outputproof-server is not recognized, keep using the python -m outputproof_server.cli form above.

The default database path is ~/.outputproof/outputproof-server.db. You can also set OUTPUTPROOF_SERVER_DB.

See The Dashboard Populate

The dashboard reads from the server API and SQLite database. To create a real dashboard record from the SDK, use this copy/paste check.

In one PowerShell window:

cd C:\workspace\ai-agent-output-verification\outputproof-server
python -m outputproof_server.cli --port 8080

Open http://127.0.0.1:8080. An empty dashboard means the server is running but no verification records have been written yet.

In a second PowerShell window:

cd C:\workspace\ai-agent-output-verification\outputproof-sdk

@'
assertions:
  - type: output_matches
    pattern: authenticated
'@ | Set-Content demo-assertions.yaml

$env:OUTPUTPROOF_SERVER_URL = "http://127.0.0.1:8080"
python -m outputproof.cli.main verify --agent-id demo-agent --prompt "Create auth" --output "authenticated" -a demo-assertions.yaml

Refresh http://127.0.0.1:8080; the summary cards, recent verifications table, agent reliability table, and team reliability leaderboard should update.

Expected dashboard change:

Total Verifications: 1
Pass Rate: 100%
Recent Verifications: demo-agent / PASS
Agent Reliability: demo-agent
Team Reliability Leaderboard: demo-agent

You can also insert a dashboard record directly through the API:

$body = @{
  request_id = "manual-demo-1"
  agent_id = "demo-agent"
  verdict = "PASS"
  confidence_score = 1.0
  assertion_results = @()
} | ConvertTo-Json -Depth 5

Invoke-RestMethod `
  -Method Post `
  -Uri "http://127.0.0.1:8080/api/verifications" `
  -ContentType "application/json" `
  -Body $body

License

outputproof-server is licensed under the Business Source License 1.1. It converts to Apache License 2.0 on the Change Date listed in LICENSE.

The Additional Use Grant allows internal self-hosted production use for small teams while requiring a commercial license for hosted service, resale, OEM, external redistribution, team governance, SSO, SLA, audit export, and cloud sync offerings.