nanobot quick reference
This is not a second tutorial. It is an entry point for returning quickly to evidence after completing the core learning path. Start with the table that matches your question, then return to the relevant chapter for context.
Find an entry point by question
| What are you asking? | Check first | Then read |
|---|---|---|
| How does one term differ from a nearby concept? | Glossary | The corresponding topic chapter |
| Which file, class, function, and test define a behavior? | Source and test index | Source tree and reading order |
| What does a configuration field or environment variable control? | Configuration and environment-variable index | Official configuration documentation |
| Which layer owns a message, event, or streamed value? | Event and protocol index | Streaming, errors, and stop |
| nanobot is not behaving as expected. Where should I look first? | Troubleshooting table | Minimal troubleshooting flow |
Usage principles
- Tables locate evidence; they do not replace understanding the call chain.
- All source paths are relative to the current
example/nanobot/code-EN. - Consult the official documentation first for complete public lists of configuration and protocol fields.
- Metadata fields beginning with an underscore are internal coordination conventions, not automatically stable public APIs.
- After locating source, read at least one adjacent test to confirm observable behavior.
These quick-reference tables describe the current source snapshot
This section is grounded in the source and tests in the current workspace. It makes no claim about other nanobot versions.
Suggested search commands
Set-Location "example\nanobot\code-EN"
# Find definitions
rg -n "class AgentLoop|class AgentRunner|class SessionManager" nanobot
# Find callers
rg -n "process_direct\\(|run_turn\\(|publish_inbound\\(" nanobot tests
# Let test names expose behavior
rg -n "cancel|fallback|stream|session" testsReturn to Learning roadmap.