TraceProof Runtime for OpenClaw

TraceProof adds provenance-first verification to real OpenClaw conversations.

This package enables:

• one TraceProof conversation trace per session
• verified inbound message proofs
• verified outbound message proofs

Register a free developer account at traceproof.org to create an Agent and credential for testing.

Important: this package enables TraceProof verification, but it does not automatically render proof state in the visible chat UI. Showing badges, verify links, QR codes, or proof summaries in a channel or app UI is the responsibility of the developer integrating OpenClaw.

What this release does

• Bundles the TraceProof runtime plugin for OpenClaw
• Creates one TraceProof conversation trace on the first message in a session
• Issues and verifies inbound message proofs (message:received)
• Issues and verifies outbound message proofs (message:sent)
• Stores per-session state in the agent workspace for trace continuity and dedupe
• Dedupes duplicate Telegram outbound events using messageId
• Dedupes duplicate inbound events using digest + short time window
• Includes bundled TraceProof grounding files for use in the OpenClaw workspace

What this release does not do

• It does not force visible per-channel UI banners or badges
• It does not fetch and render QR image binaries in chat
• It does not replace channel-specific UX logic inside OpenClaw core
• It does not automatically make the assistant knowledgeable about TraceProof unless the bundled workspace grounding is also copied into the active workspace

This release is the stable core build.

Treat Telegram as the reference proofing channel, not the reference UX-banner implementation.

UI / channel display note

This package enables TraceProof in OpenClaw conversations, but it does not force TraceProof status, badges, verify links, or QR codes to appear in the visible chat UI.

In practice, this means:

• the plugin creates the conversation trace
• the plugin issues and verifies inbound/outbound message proofs
• the developer decides how to surface verify URLs, public references, QR codes, or proof status in their own channel or app UI

Package contents

• index.js — runtime plugin
• openclaw.plugin.json — manifest + config schema
• bootstrap/AGENTS.md — workspace grounding file
• skills/traceproof/SKILL.md — public TraceProof skill
• openclaw.config.template.json — config template with placeholders
• RELEASE-NOTES.md — release scope and limits

Installation

Installation has two parts:

• install the runtime plugin
• copy the bundled TraceProof grounding into the active OpenClaw workspace

1) Install the TraceProof runtime plugin

From the package folder:

openclaw plugins install /absolute/path/to/traceproof-openclaw
openclaw plugins enable traceproof-runtime
openclaw gateway restart

This installs the core runtime that:

• creates a TraceProof conversation trace
• issues and verifies inbound/outbound message proofs
• stores per-session state in <workspace>/.traceproof-runtime/sessions.json

2) Install the TraceProof workspace grounding

For the best and most predictable TraceProof product answers in chat, also copy the bundled grounding into the active OpenClaw workspace.

Windows (PowerShell)

Copy-Item ".\bootstrap\AGENTS.md" "$HOME\.openclaw\workspace\AGENTS.md" -Force
New-Item -ItemType Directory -Path "$HOME\.openclaw\workspace\skills\traceproof" -Force
Copy-Item ".\skills\traceproof\SKILL.md" "$HOME\.openclaw\workspace\skills\traceproof\SKILL.md" -Force
openclaw gateway restart

macOS / Linux

cp ./bootstrap/AGENTS.md ~/.openclaw/workspace/AGENTS.md
mkdir -p ~/.openclaw/workspace/skills/traceproof
cp ./skills/traceproof/SKILL.md ~/.openclaw/workspace/skills/traceproof/SKILL.md
openclaw gateway restart

This step is recommended because:

• AGENTS.md in the workspace root provides the TraceProof operating guidance
• skills/traceproof/SKILL.md in the workspace makes the TraceProof skill available from the highest-precedence location

Configuration

Create a free developer account at traceproof.org, then create:

• a TraceProof Agent
• a TraceProof credential

Add this block to ~/.openclaw/openclaw.json and replace the placeholders:

{
  "plugins": {
    "entries": {
      "traceproof-runtime": {
        "enabled": true,
        "config": {
          "apiBaseUrl": "https://api.traceproof.org",
          "agentId": "<YOUR_TRACEPROOF_AGENT_ID>",
          "credentialKey": "<YOUR_TRACEPROOF_CREDENTIAL_KEY>",
          "channel": "chat",
          "sourceSystem": "openclaw",
          "originLabel": "openclaw-chat",
          "qrFormat": "svg",
          "verifyProofs": true,
          "debug": false
        }
      }
    }
  }
}

First test

• Start or open a real OpenClaw conversation
• Send a message such as:

• What is TraceProof?
• tell me about traceproof

• Wait for the assistant reply
• Check the session state file

Example check

macOS / Linux

cat ~/.openclaw/workspace/.traceproof-runtime/sessions.json

Windows (PowerShell)

type $HOME\.openclaw\workspace\.traceproof-runtime\sessions.json

You should see:

• one conversation trace
• publicReference
• verifyUrl
• one IN proof
• one OUT proof once the assistant replies

See samples/sessions.json.sample for an example.

State path

Per-session state is written to:

<workspace>/.traceproof-runtime/sessions.json

Recommended public positioning

Describe this package as:

TraceProof for OpenClaw — core verification runtime

That accurately matches the stable behavior in this release.

Summary

This package gives OpenClaw:

• a TraceProof conversation trace
• verified inbound proofs
• verified outbound proofs
• persistent per-session verification state

It does not define how proof state should look in a channel UI. That part is left to the developer integrating OpenClaw into their own user experience.