skills.rest
sandgardenhq
sandgardenhqOFFICIAL
Explore all skills by this author →
GitHub

session-control

Start, stop, and steer AI agent sessions in sgai workspaces.

130|8|Updated Jan 23, 2026
Search Tags:#api#ai#workspace#session#control#sgai
One-click install
npx skills add https://github.com/sandgardenhq/sgai --skill session-control-sandgardenhq
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: session-control
Source: https://github.com/sandgardenhq/sgai/tree/main/docs/sgai-skills/session-control
Command: npx skills add https://github.com/sandgardenhq/sgai --skill session-control-sandgardenhq

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Managing AI agent sessions in sgai workspaces is cumbersome; users need a reliable way to start, stop, and steer running sessions without manual API calls.

Core Features & Use Cases

  • Start Sessions: Launch a new agentic session in a workspace, choosing between brainstorming (interactive) and self‑drive (autonomous) modes.
  • Stop Sessions: Gracefully halt a running session, freeing resources and allowing safe revisions.
  • Steer Sessions: Inject new instructions into an ongoing session to adjust focus or priorities without restarting.
  • Workflow Insight: Retrieve the current workflow diagram as an SVG and monitor detailed session state.

Quick Start

Ask the skill to start a self‑drive session for workspace “my‑project”.

Frequently Asked Questions about session-control

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

FAQPage Schema
How do I start an AI agent session in an sgai workspace?▼

Start an AI agent session by requesting a launch in the target sgai workspace, then select either interactive brainstorming mode or autonomous self-drive mode for execution.

Can I inject new instructions into a running AI session without restarting?▼

Yes, you can steer a running AI session by injecting new instructions to adjust focus or priorities on the fly, avoiding the need to halt and restart the workflow.

Do I need a configured GOAL.md file to control sgai workspace sessions?▼

Yes, applying session control to an sgai workspace requires a configured GOAL.md file present in the workspace, alongside a running sgai server exposing REST API endpoints.

What is the best way to stop a running AI agent session gracefully?▼

The best way to stop a running AI agent session is to issue a halt command, which gracefully terminates the process, frees system resources, and allows safe workspace revisions.

How can I monitor the state of an ongoing AI session?▼

You can monitor ongoing AI session state by retrieving detailed workflow information and visualizing the current workflow diagram as an SVG file through the provided session control endpoints.

Related Skills

Explore More Software Engineering Skills→
sandgardenhqsandgardenhq
130

compose

Generate GOAL.md workflow specifications via the sgai API compose wizard.

Official
Intermediate
deskagent-aideskagent-ai
1

ui-agent-lifecycle

Explain DeskAgent UI components and agent lifecycle workflows.

Official
Advanced
sandgardenhqsandgardenhq
130

knowledge

List agents, skills, and code snippets from a SGai workspace.

Official
Intermediate
ucirelloucirello
2

session-control

Start, stop, and steer AI agent sessions in Sgai workspaces.

Community
Intermediate