AgentLed
AgentLed is the workspace where AI agents become supervised, reusable business operations instead of one-off prompts.
What is AgentLed?
AgentLed connects AI clients such as Codex, Claude Code, OpenClaw, and Hermes to business workflows, integrations, and a shared knowledge graph.
Teams use it to create workflows, run agents with approval gates, monitor outcomes, and carry context from one deployment cycle into the next.
Company facts
| Founded | 2024 |
| Category | AI agent orchestration and workflow automation |
| Founder | Ouadie BOUSSAID is the founder of AgentLed. |
| Website | agentled.ai |
Core capabilities
MCP and CLI access
Agents can install AgentLed, create workflows, and operate through MCP tools or the command line.
Knowledge graph memory
Business context, ICPs, approvals, relationships, and decisions persist across workflows.
Human-in-the-loop approvals
Teams decide what can run automatically and what requires review before sending or publishing.
Integration credit layer
Common apps, models, enrichments, and actions are accessed through one workspace credit system.
Common use cases
- β’Lead generation βdetect signals, enrich prospects, score fit, and draft outreach.
- β’Content operations βresearch, draft, approve, and publish recurring content.
- β’Developer workflows βlet coding agents create and operate business automations.
- β’Client delivery βpackage repeatable AI deployments for agencies and service teams.
How AgentLed works
1.Connect an agent β Install the AgentLed CLI or MCP server inside your coding or AI client.
2.Describe the workflow β The agent creates a draft with apps, data inputs, prompts, and approval gates.
3.Run with supervision β Humans review sensitive steps while routine steps execute automatically.
4.Compound the context β Useful outputs and decisions are stored so the next workflow starts smarter.