AI Agent Engineer

PROJECT DIRECTORY55 documents

nanobot source learning roadmap

↳Project overview
  1. Prerequisites4
    1. Prerequisite map
    2. Python and asyncio minimum
    3. LLM messages, streaming, and tool calling
    4. Source reading and debugging minimum
  2. Runtime And Source Map5
    1. Runtime and source map
    2. Complete your first request with DeepSeek
    3. Five entry points and their runtime boundaries
    4. Source tree and reading order
    5. Minimal troubleshooting flow
  3. One Turn8
    1. One Turn learning path
    2. Messages, Session keys, and entry points
    3. AgentLoop state machine
    4. ContextBuilder and model context
    5. AgentRunner's model–Tool loop
    6. Tool registration, validation, and execution
    7. Streaming, errors, and stop
    8. Persistence, response, and Runtime Events
  4. State And Long Running Work6
    1. State and long-running work
    2. Session files and replay boundaries
    3. Compaction, Memory, and Dream
    4. Goals and Continuations
    5. Cron and Heartbeat
    6. Subagent lifecycle
  5. Runtime Surfaces5
    1. Runtime surfaces
    2. Channels and Gateway
    3. HTTP API and Python SDK
    4. WebSocket and WebUI
    5. Provider resolution and fallback
  6. Engineering And Security6
    1. Engineering and security learning path
    2. asyncio concurrency, cancellation, and cleanup
    3. Errors, retries, and data consistency
    4. Logging, Runtime Events, and testing
    5. Workspace, shell, and SSRF security boundaries
    6. Packaging, WebUI build, and cross-platform behavior
  7. Extension Development7
    1. Extension development
    2. Add a Tool
    3. Add a Skill
    4. Integrate MCP
    5. Add a Provider
    6. Add a Channel
    7. Command, API, SDK, and WebUI extension boundaries
  8. Source Practice7
    1. Source practice
    2. Trace a Tool-free Turn
    3. Trace a Tool Call and Session
    4. Reproduce streaming, cancellation, and retry
    5. Write a Workspace Skill
    6. Build a read-only Tool plugin
    7. Capstone extension design review
  9. Quick Reference6
    1. nanobot quick reference
    2. nanobot and Agent development glossary
    3. nanobot source and test index
    4. nanobot configuration and environment-variable index
    5. nanobot event and protocol index
    6. nanobot troubleshooting table
AAEAI Agent Engineer
RoadmapLandmark readingsExample projectsResources中文
PERSONAL PROGRESS

Account and learning progress

Account service is not configured yet. Public courses remain fully readable.

Sign in to save a role preference and course-level progress across devices.

Signed in. Learning data is private to this account by default.

Delete account

This permanently deletes the account, role preference, and all course progress.

Privacy notice and data controls

Home

❯

examples

❯

nanobot source learning roadmap

❯

03 one turn

Folder: examples/nanobot/03-one-turn

8 items under this folder.

  • Aug 31, 2026

    Messages, Session keys, and entry points

    • nanobot
    • messagebus
    • session
  • Aug 31, 2026

    AgentLoop state machine

    • nanobot
    • agentloop
    • state-machine
  • Aug 31, 2026

    ContextBuilder and model context

    • nanobot
    • contextbuilder
    • context-engineering
  • Aug 31, 2026

    AgentRunner's model–Tool loop

    • nanobot
    • agentrunner
    • toolcalling
  • Aug 31, 2026

    Tool registration, validation, and execution

    • nanobot
    • tool
    • toolregistry
  • Aug 31, 2026

    Streaming, errors, and stop

    • nanobot
    • streaming
    • error-handling
    • stopreason
  • Aug 31, 2026

    Persistence, response, and Runtime Events

    • nanobot
    • session
    • runtimeevent
    • outboundmessage
  • Aug 31, 2026

    One Turn learning path

    • nanobot
    • agentloop
    • agentrunner
    • turn

8 items under this folder.

  • Aug 31, 2026

    Messages, Session keys, and entry points

    • nanobot
    • messagebus
    • session
  • Aug 31, 2026

    AgentLoop state machine

    • nanobot
    • agentloop
    • state-machine
  • Aug 31, 2026

    ContextBuilder and model context

    • nanobot
    • contextbuilder
    • context-engineering
  • Aug 31, 2026

    AgentRunner's model–Tool loop

    • nanobot
    • agentrunner
    • toolcalling
  • Aug 31, 2026

    Tool registration, validation, and execution

    • nanobot
    • tool
    • toolregistry
  • Aug 31, 2026

    Streaming, errors, and stop

    • nanobot
    • streaming
    • error-handling
    • stopreason
  • Aug 31, 2026

    Persistence, response, and Runtime Events

    • nanobot
    • session
    • runtimeevent
    • outboundmessage
  • Aug 31, 2026

    One Turn learning path

    • nanobot
    • agentloop
    • agentrunner
    • turn

Created with Quartz © 2026