add observer runtime model

Instrument command, payload, update, scene, runner, polling, and policy flows
Add observer regression coverage and mark observability backlog item done
This commit is contained in:
2026-03-31 13:50:21 +03:00
parent 140f3397b2
commit a818174fbf
18 changed files with 1255 additions and 62 deletions
+4 -4
View File
@@ -12,16 +12,16 @@ Russian page:
Current priority split:
- `Priority 1`: observability model.
- `Priority 2`: service layer and dependency graph model.
- `Partial`: webhook runtime model, plugin composition contract.
Completed former high-priority items:
- `[v1.0.0-rc.13] Observability model`: added first-class `Observer` events for update, command, payload, scene, policy, runner, polling, and centralized error flows, with safe event dispatch and regression coverage for the new runtime hooks.
- `[v1.0.0-rc.13] Authorization and policy model`: added first-class `Policy[T]`, middleware integration through `RequirePolicy(...)`, plugin and bot policy registration helpers, built-in Telegram-aware policies, and composable `AllPolicies(...)`, `AnyPolicy(...)`, and `NotPolicy(...)` helpers with regression coverage.
- `[v1.0.0-rc.13] Update schema contract`: documented and tested the normalized `MsgContext` update-routing contract, including routing categories and per-update field guarantees.
- `[v1.0.0-rc.13] User-facing vs internal error model`: added explicit user-visible vs internal-only error markers and updated centralized handler error routing accordingly.
- `[v1.0.0-rc.13] Configuration freeze model`: formalized bot configuration freeze after first run, documented lifecycle commit points, and added regression coverage for ignored late mutations.
- `1. Conversation / Scene Model`: completed in `v1.0.0-rc.12`.
- `2. Typed Handler Input Model`: completed in `v1.0.0-rc.12`.
- `3. Request Context / Cancellation Model`: completed in `v1.0.0-rc.12`.
- `[v1.0.0-rc.12] Conversation / Scene Model`.
- `[v1.0.0-rc.12] Typed Handler Input Model`.
- `[v1.0.0-rc.12] Request Context / Cancellation Model`.