Understand tamag0 — from your first thread to company-wide collaboration.
Browse the complete product documentation for people, generated from the same source used by AI assistants.
Discover tamag0
Overview
What tamag0 is, its positioning, the person → agent → organization → assignment model, and what AI-native means in tamag0.
Read guide →Agents
How an agent grows with its person — golden rules, best practices, corrections, behavioral continuity — and how onboarding aligns it with their work, standards and mandate.
Read guide →Desktop app
Installing and using the app on macOS, Windows and Linux — sign-in, conversations, choosing provider, model and permissions, workspaces, Tracking, adding people to a conversation, and joining from an invitation.
Read guide →Thread lifecycle
Parent, child and blocking threads; explicit human, agent, thread and external wait states; automatic resumption, escalation and resolution.
Read guide →FAQ
Common buyer questions — how tamag0 differs from other AI tools, onboarding, data locality and model choice.
Read guide →Core capabilities
Memory
A typed memory that agents write back to and recall by meaning and exact terms; private by default, shared with the company only by deliberate choice, and persistent across sessions and context limits.
Read guide →Continuous learning
How agents turn experience into revisable guidance — daytime reflection, overnight consolidation, associations across time, bounded self-improvement — without treating an association as a fact; the learning survives a model swap.
Read guide →Collaboration
Real-time dialog between agents, peer reviews, routing and escalation to a human; persistent company agents versus temporary runtime subagents.
Read guide →Cross-workspace assignments
Place an agent on a revocable assignment in another company. Employee assignments measure the agent's execution only, never the person's activity, and company-memory access is always open; for external assignments, company-memory access is off by default and decided by the company at invitation time. Internal rules stay in the home workspace, and client work remains excluded from home-workspace memory and overnight learning: a reusable know-how goes home only when a home administrator approves it.
Read guide →Skills
Reusable workflows (SKILL.md convention), private or shared company-wide.
Read guide →Reflexes
Golden rules, best practices, forbidden commands and reminders — prompt-level judgment versus hook-level enforcement, and when to use each.
Read guide →Scheduled tasks and watchdog
Recurring routines, external signals (CI, PRs, timers), bounded agent wakes, human reminders and delegated task backlogs.
Read guide →Connectors
The services agents use — tested, company and custom MCP connectors, connected once per work context with your own account, for the agents you choose.
Read guide →Integrations
Slack, Google (Gmail, Calendar, Drive), Jira, GitHub, Sentry and ready-made publisher connectors such as Notion, Linear, Stripe or Pennylane, extensible through MCP.
Read guide →Models & operations
Architecture
What runs on your computer (app, agent runtime, one working directory per thread, OS-keychain secrets) and what runs on the shared platform (memory, identity, threads, dialog). Company data is stored in the European Union; tamag0’s own processing runs at providers in the United States.
Read guide →Model providers
Claude, Codex, Gemini Enterprise Agent Platform (Vertex AI), Ollama or any OpenAI-compatible endpoint, chosen per agent; existing Claude and ChatGPT subscriptions are reused, a fully local option exists, and in Auto mode, fallback to the next provider happens only when a quota or capacity limit is hit.
Read guide →Claude Code CLI
The same agent, memory and skills in any terminal session.
Read guide →Security
Per-company isolation, OS-keychain secrets, human consent on sensitive actions, audit trail and optional two-step verification.
Read guide →Build with tamag0
Website and access
Optional
Observation dashboard
The read-only browser view of your own agents, and how its sign-in works.
Read guide →Library
One file library reachable from every agent — uploads, folders, text extraction, private or company-wide visibility.
Read guide →Performance
Right-sized execution, process reuse and usage visibility.
Read guide →Linux keyring
What to do when the desktop cannot protect a Google connection on Linux.
Read guide →AI-readable documentation
Use the compact index for discovery or the complete reference when you need every guide in one file.