Lobster

Orchestrate deterministic multi-step workflows with human approval gates.

202|36|Updated Jan 27, 2026
One-click install
npx skills add https://github.com/adoresever/AGI_Ananas --skill lobster-adoresever
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Lobster
Source: https://github.com/adoresever/AGI_Ananas/tree/main/26.2.21openclaw-viking/extensions/lobster
Command: npx skills add https://github.com/adoresever/AGI_Ananas --skill lobster-adoresever

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Lobster provides a safe, deterministic way to run multi-step workflows that may require human approvals before actions execute, ensuring repeatability, auditing, and easy resumption after interruptions.

Core Features & Use Cases

  • Deterministic multi-step pipelines with visible outcomes and auditable history.
  • Approval gates that pause execution until a human confirms before actions execute.
  • Resumable workflows that continue from a saved state or token after a pause or failure.
  • Tool orchestration that coordinates multiple commands into a single pipeline.

Use cases include triaging emails, monitoring systems, syncing data across services, and orchestrating complex operations with checkpoints.

Quick Start

Run a multi-step workflow with approval checkpoints by providing a pipeline to execute and awaiting your approval to resume.

Frequently Asked Questions about Lobster

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run deterministic multi-step workflows with human approval gates?

Deterministic multi-step workflows with human approval gates execute visible pipeline stages and pause operations until a human confirms the action. This ensures repeatability and provides an auditable history of the coordinated tool operations before proceeding.

Can I resume a multi-step pipeline after an interruption or failure?

Resumable multi-step pipelines save their execution state or token after a pause or failure. You can resume the workflow from this saved checkpoint, ensuring complex operations continue reliably without restarting the entire tool orchestration process from the beginning.

How do I orchestrate multiple tools safely into a single automated pipeline?

Tool orchestration safely coordinates multiple commands into a single automated pipeline by validating inputs and running subprocesses securely. This enforces security and reliability while managing multi-tool operations across teams for tasks like data synchronization.

What is the best way to automate triage and monitoring workflows with required approvals?

Automating triage and monitoring workflows with required approvals involves applying deterministic pipelines with visible outcomes. The workflow pauses execution at approval gates, ensuring a human confirms before actions execute, which maintains security and auditability.

Do I need specific dependencies to synchronize data across services with approval checkpoints?

Synchronizing data across services with approval checkpoints requires no specific dependencies. The workflow validates inputs and runs subprocesses safely within your existing environment, coordinating multiple commands into a single resumable pipeline without external tool prerequisites.