Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.bossmode.ing/llms.txt

Use this file to discover all available pages before exploring further.

🚧 Expanding soon — full OpenClaw integration documentation is in progress.

​
BossMode + OpenClaw

OpenClaw is an agent harness for running autonomous AI workers. The BossMode-OpenClaw integration layers governance on top of OpenClaw runs — standing orders are enforced, every move is logged, and the Daily Report captures what happened.

​
Integration overview

When BossMode governance is enabled in OpenClaw:
  • Standing orders are injected as system-level constraints on every agent run
  • Moves that exceed standing order thresholds are held for approval before execution
  • All agent activity is logged to the BossMode audit feed
  • The consigliere reviews OpenClaw output and surfaces it in the Daily Report

​
Setup

Install the BossMode OpenClaw plugin: 
npm i @acelabs/bossmode-openclaw-plugin
Configure in your OpenClaw .clawrc.json: 
{
  "plugins": ["@acelabs/bossmode-openclaw-plugin"],
  "bossmode": {
    "token": "${BOSSMODE_API_TOKEN}",
    "governance": true,
    "injectStandingOrders": true,
    "logAllMoves": true,
    "requireApprovalDefault": true
  }
}

​
Usage

With governance enabled, your OpenClaw agents automatically operate within BossMode’s rules. No code changes needed in your existing agent workflows. To issue a BossMode-aware directive that runs via OpenClaw: 
curl -sX POST https://bossmode.ing/api/pro/directives \
  -H "x-bossmode-token: $BOSSMODE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "directive": "Audit the current content library and identify gaps against the quarterly content plan.",
    "capo": "capo_marketing",
    "runtime": "openclaw",
    "requiresApproval": true
  }'
The "runtime": "openclaw" field routes execution through your connected OpenClaw instance.

​
Governance enforcement

BossMode enforces governance at two levels in OpenClaw:
  1. Pre-execution — Guardian checks standing orders before the agent run begins
  2. Post-execution — Output is reviewed against standing orders before being marked complete
If a standing order violation is detected post-execution, the output is held in the approval queue even if requiresApproval was set to false.

​
Requirements

  • OpenClaw v0.6.0 or higher
  • BossMode Pro account
  • BOSSMODE_API_TOKEN set in your environment