← Back to Skills Marketplace
vanhuelsing

Flowclaw

by vanhuelsing · GitHub ↗ · v1.1.3 · MIT-0
cross-platform ⚠ suspicious
154
Downloads
0
Stars
0
Active Installs
5
Versions
Install in OpenClaw
/install flowclaw-orchestrator
Description
YAML-driven workflow orchestrator for AI agent teams with human-in-the-loop approval gates. Includes optional Notion, n8n, and Discord integrations.
README (SKILL.md)

FlowClaw

YAML-driven workflow orchestrator for OpenClaw agent teams. Connects Notion → n8n → agents, with approval gates so nothing runs without your go-ahead.

What It Does

FlowClaw is a workflow execution engine that:

  1. Receives task triggers from n8n (via Notion polling)
  2. Loads the appropriate YAML workflow definition
  3. Executes each step by dispatching to specialized AI agents
  4. Pauses at approval gates and waits for human sign-off
  5. Reports progress via Discord notifications

Requirements

  • Python 3.8+
  • OpenClaw with configured agents
  • n8n (optional — only needed for n8n-triggered workflows)
  • Notion workspace with task database (optional — only needed for Notion sync)

Setup

  1. Copy config/example.env to .env and fill in your API keys
  2. Install: pip3 install -r src/requirements.txt
  3. Start: python3 src/workflow-executor.py
  4. Import src/n8n-workflow.json into your n8n instance
  5. Update n8n credential/ID placeholders with your values — see INTEGRATION-STEPS.md

Configuration

All configuration is via environment variables. See config/example.env for the full list.

Key variables:

  • WORKFLOW_EXECUTOR_API_KEY — API key for authenticating requests
  • NOTION_API_KEY — Notion integration token (starts with secret_...)
  • DISCORD_BOT_TOKEN — Discord bot token for notifications (optional)
  • PORT — Server port (default: 8765)
  • MAX_WORKERS — Gunicorn worker count (default: 4, recommended: 2× CPU cores)
Usage Guidance
FlowClaw appears to be what it says: a local workflow executor that wires Notion/n8n/Discord into agent-driven YAML workflows. Before installing or running it in a sensitive environment, consider the following: 1) Review any workflow YAMLs and QA scripts before enabling them — QA scripts are executed as Python and are not sandboxed (they can read the FS and make network calls). 2) Keep the service bound to 127.0.0.1 (the default) and use a strong WORKFLOW_EXECUTOR_API_KEY; if you change HOST to 0.0.0.0, place a reverse proxy + auth in front of it. 3) Do not set FLOWCLAW_LOAD_OPENCLAW_CONFIG=true unless you intend to grant FlowClaw access to all credentials in ~/.openclaw/openclaw.json. 4) If you enable deploy steps that call external CLIs (e.g., vercel), be aware those CLIs run with the process environment and may use credentials available there. 5) Run in an isolated service account or container and review startup/service templates before enabling systemd/LaunchAgent. These are operational risks intrinsic to a workflow orchestrator, not indications of misbehavior by the skill itself.
Capability Analysis
Type: OpenClaw Skill Name: flowclaw-orchestrator Version: 1.1.3 FlowClaw is a sophisticated workflow orchestrator that manages AI agent tasks through YAML definitions, featuring integrations with Notion, Discord, and n8n. The application in `src/workflow-executor.py` possesses high-risk capabilities, including the ability to execute shell commands via `subprocess` (e.g., `openclaw` and `vercel` CLIs) and run Python-based QA scripts. Although the codebase is professionally authored with extensive security measures—such as strict input validation using regex, path traversal prevention via the `_safe_path` helper, and a robust security test suite in `tests/test_security.py`—the inherent power to execute code and manage sensitive API keys across multiple platforms makes it a high-risk component. Per the provided guidelines, these risky capabilities, while plausibly needed for the stated purpose, necessitate a suspicious classification.
Capability Assessment
✓ Purpose & Capability
Name/description, required env var (WORKFLOW_EXECUTOR_API_KEY), optional Notion/Discord/n8n envs, scripts, and Python app match the claimed purpose of a workflow executor and integrations. The included files (executor, startup script, n8n workflow, sample YAML workflows) are proportionate to the stated functionality.
ℹ Instruction Scope
SKILL.md keeps instructions focused (install deps, copy .env, run the executor, import n8n workflow). It explicitly documents important behaviors and security boundaries (Content-Type checks, auth, local bind default, credential isolation opt-in). Notable items to be aware of: (1) FlowClaw executes user-supplied QA Python scripts (allowed but explicitly unsandboxed), (2) deploy steps may invoke external CLIs (e.g., vercel) which run with the service's environment, and (3) there are operational instructions to run as a system service (LaunchAgent/systemd) which is expected for a long-lived executor.
✓ Install Mechanism
No remote download or opaque installer; this is an instruction-only skill with source included. Dependencies are standard Python packages listed in requirements.txt and the SKILL.md tells the operator to run pip install -r src/requirements.txt. No external URLs or archives are fetched by an automated install spec in the skill metadata.
ℹ Credentials
Only a single required env var (WORKFLOW_EXECUTOR_API_KEY) is declared; other tokens (NOTION_API_KEY, DISCORD_BOT_TOKEN, N8N_API_KEY, OPENCLAW_GATEWAY_TOKEN, etc.) are optional and justified by the optional integrations. The README and config explicitly document an opt-in flag (FLOWCLAW_LOAD_OPENCLAW_CONFIG) that, if enabled, will read ~/.openclaw/openclaw.json and thereby expose other OpenClaw credentials — that is an explicit, documented escalation and is off by default.
✓ Persistence & Privilege
The skill is not always-enabled and does not request elevated autonomous platform privileges. It includes startup/service templates and scripts to run as a persistent service (LaunchAgent/systemd), which is normal for this type of component. No attempt to modify other skills' configurations or system-wide agent settings was observed.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install flowclaw-orchestrator
  3. After installation, invoke the skill by name or use /flowclaw-orchestrator
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.1.3
Final QA fixes: datetime handling, systemd integration, schema path resolution, SKILL.md alignment. 81/81 tests passing.
v1.0.3
Security: credential isolation opt-in, agent validation, path checks
v1.0.2
Security: bind 127.0.0.1 default, credential docs, env var declarations
v1.0.1
Fix: Notion, n8n, and Discord are optional integrations, not requirements.
v1.0.0
Initial release
Metadata
Slug flowclaw-orchestrator
Version 1.1.3
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 5
Frequently Asked Questions

What is Flowclaw?

YAML-driven workflow orchestrator for AI agent teams with human-in-the-loop approval gates. Includes optional Notion, n8n, and Discord integrations. It is an AI Agent Skill for Claude Code / OpenClaw, with 154 downloads so far.

How do I install Flowclaw?

Run "/install flowclaw-orchestrator" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Flowclaw free?

Yes, Flowclaw is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Flowclaw support?

Flowclaw is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Flowclaw?

It is built and maintained by vanhuelsing (@vanhuelsing); the current version is v1.1.3.

More Skills
self-improving agent
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
⬇ 463283 · by pskoett
Skill Vetter
Security-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns.
⬇ 259410 · by spclaudehome
Polymarket
Query Polymarket prediction markets. Check odds, find trending markets, search events, track price movements.
⬇ 232503 · by joelchance
Self-Improving + Proactive Agent
Self-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when...
⬇ 200423 · by ivangdavila
Github
Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.
⬇ 191115 · by steipete
ontology
Typed knowledge graph for structured agent memory and composable skills. Use when creating/querying entities (Person, Project, Task, Event, Document), linkin...
⬇ 190156 · by oswalpalash

💬 Comments