Authoritative docs and implementation
- Config base class docs:
docs/config-base.md - SystemConfig:
core/system-Core/commonconfig/system.js - Config schemas:
core/system-Core/commonconfig/*.js
Configuration path rules (core)
Authoritative constants:src/infrastructure/config/config-constants.js
| Category | Storage path | Config name (example) |
|---|---|---|
| Global | data/server_bots/<name>.yaml |
agt、device、monitor、notice、mongodb、redis |
| Along with port | data/server_bots/{port}/<name>.yaml |
server、chatbot、group、aistream |
| Factory LLM | data/server_bots/{port}/<name>.yaml |
openai_llm、volcengine_llm、gemini_compat_llm etc. (see FACTORY_CONFIG_PATTERNS) |
- Default templates:
config/default_config/<name>.yaml(Cfgautomatically copied to the target path when missing) - Schema sources:
- Standalone files:
core/system-Core/commonconfig/<name>.js(same name as the YAML, e.g.openai_llm.js) - Aggregated in
system.js:agt、server、chatbot、device、group、notice、redis、mongodb、monitor、aistreametc. (no separate commonconfig file) - Code-side only:
tools.js(no corresponding default YAML)
- Standalone files:
Change checklist (must check when making config-related changes)
config/default_config/<name>.yaml: whether the default template fields are completecore/system-Core/commonconfig/<name>.js: whether schema fields correspond 1:1, with reasonable defaults/enums- Whether client/factory code actually consumes these fields (avoid "schema written but unused")
- If system-Core's
.gitignorehas a whitelist: new commonconfig files must be added to the whitelist
aistream specifics
- Runtime:
data/server_bots/{port}/aistream.yaml(cfg.aistream); template:config/default_config/aistream.yaml。 - Common sections:
embedding、mcp、agentWorkspace、tools(file→ToolsStream、web.fetch→WebStream、agentBrowser→BrowserStream); desktop workflow tools are implemented indesktop.js, with notools.desktopYAML。
Authoritative entry points
- Project overview:
PROJECT_OVERVIEW.md - Code entry points:
src/andcore/corresponding subdirectories - Related docs:
docs/corresponding topic docs under
Applicable scenarios
- Need to locate the implementation path and configuration entry points of this subsystem.
- Need to quickly provide change locations and compatibility notes.
Non-applicable scenarios
- Not used to replace implementation descriptions of other subsystems.
- Do not fabricate paths or fields when evidence is lacking.
Execution steps
- First confirm the requirement falls within this skill's responsibility boundary.
- Then provide code paths, configuration paths, and key fields.
- Finally add risk points, verification steps, and regression scope.
Common pitfalls
- Giving only concepts without specific file paths.
- Not noting that code prevails when docs and code conflict.
- Ignoring consistency among configuration, Schema, and consuming code.