Agents can sign plugins, rotate credentials without losing identity, and publicly attest to behavior.
日本語の概要は準備中です。原文の説明を表示しています。
General-purpose AI workflow framework for OpenClaw. Build custom multi-step workflows for any task — dev, ops, research, content, or automation. Ships with dev workflow examples.
インストールする前に、エージェントに与えられる指示の中身を確認できます。
A framework for building custom AI workflows. Each workflow is a series of steps that execute in isolated agent sessions with full tool access. Define your workflow in JSON, invoke it naturally, and let the agents handle the execution.
What you can build:
Ships with 3 example dev workflows to show you how it works. Build your own for anything.
Natural language (easiest): Just tell your agent what you need:
Your agent reads this SKILL.md and invokes SoulFlow automatically.
Command line:
cd ~/.openclaw/workspace/soulflow
# Run a security audit
node soulflow.js run security-audit "Audit the codebase at ~/project for vulnerabilities"
# Fix a bug
node soulflow.js run bug-fix "Login returns 500 when email has uppercase letters in ~/myapp"
# Build a feature
node soulflow.js run feature-dev "Add dark mode toggle to the settings page in ~/myapp"
SoulFlow connects to your local OpenClaw gateway via WebSocket and runs each workflow step as an isolated agent session. A dedicated soulflow-worker agent is auto-created with minimal context — no memory bleed from your main agent.
Each step:
Auto-notifications (v1.1.0+): When workflows complete, SoulFlow automatically notifies the main agent session with results. No need to manually check status.
These are examples to show what's possible. Build your own for any domain.
Scan → Prioritize → Fix → Verify Development example: Reads your source files, identifies vulnerabilities by severity, applies fixes, then verifies them.
Triage → Fix → Verify Development example: Investigates the root cause by reading code, applies the fix, then verifies it didn't introduce regressions.
Plan → Implement → Review Development example: Architects the implementation plan, writes the code, then reviews for quality and correctness.
Want content workflows? Research pipelines? Deploy automation? Create your own .workflow.json — see Custom Workflows below.
node soulflow.js run <workflow> "<task>" # Run a workflow
node soulflow.js list # List available workflows
node soulflow.js runs # List past runs
node soulflow.js status [run-id] # Check run status
node soulflow.js test # Test gateway connection
The agent knows how to invoke SoulFlow for you. Just describe what you want:
Security audits:
Bug fixes:
Features:
How it works:
node soulflow.js run <workflow> "<task>"Pattern matching: The agent matches your message to workflows:
No workflow matches? Agent will ask which workflow you want or suggest creating a custom one.
You can create workflows for ANY task. Define them in JSON and place in the workflows/ directory.
Tell your agent:
"Create a SoulFlow workflow for [your use case]"
Examples:
Your agent will:
.workflow.json file to workflows/Create a .workflow.json file in the workflows/ directory:
{
"id": "my-workflow",
"name": "My Custom Workflow",
"version": 1,
"description": "What this workflow does",
"steps": [
{
"id": "step1",
"name": "First Step",
"input": "Do this thing: {{task}}",
"expects": "STATUS: done",
"maxRetries": 1
},
{
"id": "step2",
"name": "Second Step",
"input": "Now do this based on step 1:\n\n{{step1_output}}\n\nOriginal task: {{task}}",
"expects": "STATUS: done",
"maxRetries": 1
}
]
}
{{task}} — The user's original task description{{stepid_output}} — Full output from a previous step (e.g. {{scan_output}})KEY: value lines in step output become variables (e.g. ROOT_CAUSE: ... → {{root_cause}})For best results, write prompts that:
read to examine the file", "Use edit to apply the fix"soulflow-worker agent with minimal brain files~/.openclaw/workspace/.soulflow/runs/What SoulFlow does to your OpenClaw instance:
~/.openclaw/openclaw.json) to obtain the authentication token needed to connect via WebSocket~/.openclaw/openclaw.json) via config.patch to register the soulflow-worker agentsoulflow-worker) with minimal brain files (SOUL.md only, no memory/history)~/.openclaw/workspace/.soulflow/runs/ as JSON filesWhy these permissions are needed:
openclaw CLI tool)Security considerations:
Recommended practices:
.workflow.json files before running~/.openclaw/workspace/.soulflow/runs/ for workflow execution logsWhen the user requests a workflow (security audit, bug fix, feature build, etc.), you should:
Identify the workflow by matching keywords:
workflows/*.workflow.json for other optionsExtract the task description — the user's description of what they want done
Invoke SoulFlow using exec:
cd /root/.openclaw/workspace/soulflow && node soulflow.js run <workflow> "<task>"
Monitor the run — SoulFlow will output the run ID, then show progress as each step completes
Report results — When complete, relay the final status to the user
Example:
User: "Run a security audit on ~/myapp"
You: [exec] cd /root/.openclaw/workspace/soulflow && node soulflow.js run security-audit "Audit ~/myapp for vulnerabilities"
Creating workflows for users: If the user asks you to create a custom workflow:
.workflow.json file to /root/.openclaw/workspace/soulflow/workflows/See CONTRIBUTING.md for workflow design best practices.
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Agents can sign plugins, rotate credentials without losing identity, and publicly attest to behavior.
日本語の概要は準備中です。原文の説明を表示しています。
Query historical crypto market data from 0xArchive across Hyperliquid, Lighter.xyz, and HIP-3. Covers orderbooks, trades, candles, funding rates, open interest, liquidations, and data quality. Use when the user asks about crypto market data, orderbooks, trades, funding rates, or historical prices on Hyperliquid, Lighter.xyz, or HIP-3.
日本語の概要は準備中です。原文の説明を表示しています。
Calculate ROI for AI-as-a-Service (managed AI agents). Estimates cost savings, efficiency gains, and payback period for deploying AI agents across business operations. Use when evaluating whether managed AI agents make financial sense for a company.
日本語の概要は準備中です。原文の説明を表示しています。
Design, build, deploy, and operate production AI agent systems — single agents, multi-agent teams, and autonomous swarms. Complete methodology from agent architecture through orchestration, memory systems, safety guardrails, and operational excellence.
日本語の概要は準備中です。原文の説明を表示しています。
Design, build, test, document, and secure production-grade APIs. Covers the full lifecycle from schema design through deployment, monitoring, and versioning. Use when designing new APIs, reviewing existing ones, generating OpenAPI specs, building test suites, or debugging production issues.
日本語の概要は準備中です。原文の説明を表示しています。
Turn your AI agent into a business automation architect. Design, document, implement, and monitor automated workflows across sales, ops, finance, HR, and support — no n8n or Zapier required.
日本語の概要は準備中です。原文の説明を表示しています。