Create focused AI agents
Add your provider API key and model name in Settings, then create agents with roles, prompts, capability, and knowledge.
Browser based AI Agent OS
Create agents, connect social channels like Facebook, Instagram, WhatsApp, and website widgets, then run skills, workflows, leads, sessions, analytics, and logs from one operating workspace.
Create AI agents with provider keys, model names, prompts, and knowledge.
Turn website widgets, social accounts, WhatsApp, and APIs into AI entry points.
Use skills, workflows, cronjobs, logs, and usage signals to operate reliably.

Channel
WhatsApp message
Agent
Reply + lead saved
Agent plan
Skill + knowledge + workflow
Core layers
The product is organized into simple building blocks: agents, skills, workflows, knowledge, channels, sessions, leads, analytics, scheduled runs, memory consolidation, logs, and settings.
Add your provider API key and model name in Settings, then create agents with roles, prompts, capability, and knowledge.
Create skills for browser work, APIs, webhooks, files, database saves, research, and workflow actions.
Connect agents, skills, schedules, inputs, and outputs into clean multi-step workflows.
Store FAQs, product details, website notes, policies, and files so agents answer from trusted context.
Channels
Connect Facebook, Instagram, WhatsApp, Telegram, Slack, Email, Public API, and your Website Widget. Assign any connected channel to an AI agent so conversations can be answered, qualified, saved, and routed automatically.
Website widget setup
Create a website widget channel, select the default agent, then copy the script and paste it into your website.

Agent setup
The agent creation flow stays simple: go to Settings, save your provider API key and model name, then create agents from the Agents page with role, capability, prompt, and status.
Save provider key
Create agent
Add knowledge
Attach skills
Connect channel
Run workflow
Product modules
These sections match your dashboard modules so visitors can understand exactly what OpenLioness does before signing up.
Build
Create AI workers for support, sales, operations, research, and internal workflows.
3 steps
Settings → Provider → Agent
Save provider API key and model name in Settings.
Create an agent with role, capability, prompt, and status.
Assign knowledge, channels, or workflows when the agent is ready.
Actions
Reusable abilities that agents and workflows can run whenever they need action.
Reusable
Browser, API, webhook, files
Create browser automation and research skills.
Configure API calls, webhook payloads, file actions, and database saves.
Test skill input before connecting it to a workflow.
Automation
Design, save, run, and monitor multi-step automations without a bulky setup screen.
Multi-step
Agent + skill + schedule
Build workflow JSON or use saved templates.
Run once manually or connect workflows to cronjobs.
Review latest result, duration, steps, and success status.
Memory
Store product FAQs, policies, website notes, file text, and business details for your agents.
Trusted
Workspace or agent scoped
Add text, docs, product FAQs, or policy content.
Scope sources to a specific agent or the whole workspace.
Use knowledge to reduce guessing and improve answer quality.
Conversations
Review conversations from website widgets, social channels, WhatsApp, and internal agent chats.
History
Every conversation stored
Track visitor, contact, agent, activity, and channel type.
Review messages from website widget or agent chat sessions.
Use session patterns to improve prompts, knowledge, and workflows.
Capture
Capture visitor details, source context, company information, and follow-up status.
CSV
Filter and export leads
Store lead name, email, phone, source, and context.
Track status such as new, qualified, or won.
Export filtered leads for your sales or support team.
Runtime
Monitor connected runtimes, workers, widgets, workflow runners, and channel services.
Health
Runtime monitoring
Register website widget runtimes and connected services.
Track online, offline, warning, and error states.
Keep external entry points visible in one health view.
Analytics
Monitor requests, token volume, latency, provider usage, success rate, and model executions.
Insights
Requests, tokens, latency
View provider and model breakdown.
Track daily usage and average response time.
Spot failed events and performance changes quickly.
Schedule
Schedule recurring agent prompts for reports, summaries, reminders, scraping, and background intelligence.
Recurring
Daily, weekly, custom
Create scheduled runs for agents or workflows.
Use presets such as daily summaries and weekly kickoff.
Review latest run details and resume paused jobs.
Memory consolidation
Stage useful session facts, group recurring patterns, and promote high-value memories for agents.
Deep sleep
Promote useful memory
Run memory sweeps across recent sessions.
Group repeated facts and filter low-value noise.
Promote useful memories for long-term agent context.
Observability
Inspect workspace activity, automation errors, debugging events, and system actions from one clean view.
Debug
Activity and errors
Filter by debug, info, warning, or error.
Inspect source, entity, message, and metadata.
Use logs to troubleshoot workflows, instances, and skill runs.
Admin
Manage profile, workspace branding, provider keys, model names, defaults, and system controls.
Keys
Encrypted provider access
Save model provider API keys securely.
Set brand name, logo URL, accent color, and workspace controls.
Update profile, workspace, runtime, widget, and dreaming defaults.
Start with a provider key, create your agent, connect a website widget or social channel, and automate the rest with skills, workflows, cronjobs, memory, logs, and analytics.