Developer Preview
Deploy Your Character
From a local folder to an autonomous cloud agent in 5 minutes.
No Docker. No Kubernetes. Just psycheos push.
Architecture
📁
Your Folder
.md files + JSON config
⚡
PsycheOS Runtime
Local or Cloud
→
→
🌐
Real World
Email, LinkedIn, Job portals
01
Init your character folder
One command scaffolds everything. Your knowledge = .md files. Your config = JSON.
$ mkdir ~/AI/my-agent && cd ~/AI/my-agent $ psycheos init ✓ Created character.psycheos.json ✓ Created Knowledge/ folder ✓ Created .env.example Edit character.psycheos.json to define personality, skills, and schedules.
02
Add your knowledge
Drop .md files into Knowledge/. Or point to any folder on your machine. PsycheOS reads them all.
$ ls Knowledge/
resume.md # Your experience & skills
job-preferences.md # Target roles, salary, companies
cover-letter-style.md
# Or point to existing .md files:
{
"knowledge": {
"sources": ["Knowledge/*.md", "../notes/**/*.md"]
}
}03
Run locally
Test everything locally first. Chat via CLI or web UI. Tune personality until it feels right.
$ psycheos run ✓ Character loaded: "Hunter" (6 skills, 4 knowledge files) ✓ Web UI: http://localhost:3100 ✓ API: http://localhost:3100/api/chat ✓ Schedules: 4 active (next: job-hunt in 4:58) # Chat with your character locally $ psycheos chat "Find me React jobs in Singapore" 🔍 Scanning... found 5 matches 📝 Tailoring CV for each... ✉️ Applied to 3 (2 require manual portal)
04
Push to cloud
One command. Your agent is now autonomous on the cloud. Runs schedules, sends emails, applies to jobs — all without your machine.
$ psycheos push --cloud Uploading character config... ✓ Syncing knowledge (13 files, 47KB)... ✓ Activating schedules... ✓ Connecting integrations... ✓ ✓ Deployed to: https://psycheos.ai/agents/hunter ✓ Status: LIVE 🟢 ✓ Dashboard: https://psycheos.ai/dashboard/hunter Your agent is now running 24/7. Turn off your laptop. It keeps working.
05
Monitor & iterate
Check dashboard anytime. Update knowledge locally, sync to cloud. Your agent gets smarter over time.
$ psycheos status Agent: Hunter (Job Seeker Agent) Status: LIVE 🟢 (uptime: 3d 14h) ━━━━━━━━━━━━━━━━━━━━━━━━━━━ Cycles completed: 847 Jobs applied: 127 Callbacks received: 14 Interviews scheduled: 3 ━━━━━━━━━━━━━━━━━━━━━━━━━━━ Next action: job-hunt in 2:31 Last action: Applied to Stripe (Sr. Frontend, $175k) # Update knowledge — auto-syncs to cloud $ echo "- New skill: Rust + WebAssembly" >> Knowledge/resume.md $ psycheos sync ✓ Knowledge updated on cloud (1 file changed)
character.psycheos.json Reference
The complete config for an autonomous job-seeking agent:
character.psycheos.json
{
"name": "Hunter",
"role": "Job Seeker Agent",
"personality": "Relentless, strategic, works 24/7...",
"knowledge": {
"sources": ["Knowledge/*.md"],
"tags": ["Resume", "Job preferences", "Cover letters"]
},
"skills": ["web-search", "resume", "write", "auto-apply", "email", "schedule"],
"autonomous": {
"enabled": true,
"schedules": [
{
"name": "Job Hunter",
"cron": "*/5 * * * *",
"actions": ["web-search", "resume", "write", "auto-apply", "email"],
"config": {
"targets": ["senior frontend", "AI engineer"],
"salary_min": 150000,
"remote_only": true
}
}
]
},
"integrations": {
"email": { "provider": "gmail" },
"calendar": { "provider": "google-calendar" }
}
}