docs: define remaining pluginization tasks / 定義剩餘插件化 task 與下一個 migration 目標

reports/next-inline-migration-target.md

@@ -0,0 +1,36 @@
+# Next Inline Migration Target
+
+## Selected target
+
+The next inline runtime behavior to migrate should be:
+
+- **callback acknowledgement + resolved button rendering**
+
+## Why this target first
+
+- repo-side helper ownership already exists
+- runtime behavior is compact and visually testable
+- lower risk than changing stop-policy injection first
+- success is easy to verify through Telegram UI and callback acknowledgement messages
+
+## Repo-side owners already available
+
+- `src/runtime/openclaw-telegram-poc.ts`
+- `src/runtime/openclaw-telegram-bridge.ts`
+- `src/runtime/openclaw-telegram-callback-actions.ts`
+- `src/config.ts`
+- `config/reply-end-controls.json`
+
+## Migration goal
+
+Reduce the amount of hardcoded callback acknowledgement and resolved button logic that remains inline in the patched OpenClaw Telegram runtime.
+
+## Acceptance for this target
+
+After migration:
+
+- continue button still updates UI correctly
+- stop button still updates UI correctly
+- continue acknowledgement still appears
+- stop acknowledgement still appears
+- no regression in state persistence

reports/remaining-task-list.md

@@ -0,0 +1,45 @@
+# Remaining Pluginization Task List
+
+## Goal
+
+Track the remaining major work needed to move `reply-end-controls` from a successful Telegram/OpenClaw PoC toward a cleaner agent-consumable integration.
+
+## Remaining major tasks
+
+### Task A - Reduce live runtime inline patching
+
+Current runtime patching still directly owns some logic that should eventually live behind repo-side helpers/bridges.
+
+Subtasks:
+
+1. identify remaining inline logic not yet cleanly represented in repo modules
+2. pick one small inline runtime behavior to migrate next
+3. verify that migration does not break Telegram acceptance
+
+### Task B - Tighten apply / verify / rollback into a clearer consumer workflow
+
+Current scripts exist, but they still feel like engineering tools rather than a polished integration workflow.
+
+Subtasks:
+
+1. make script expectations explicit
+2. ensure output is concise and user-facing enough for another agent to follow
+3. document an end-to-end happy path
+
+### Task C - Final maturity judgment
+
+The project still needs a final written judgment on whether it is:
+
+- only suitable for further development handoff
+- suitable for direct reuse by another agent in controlled conditions
+- or ready for broader adoption
+
+Subtasks:
+
+1. define maturity criteria
+2. compare current state against those criteria
+3. write final readiness report
+
+## Immediate next move
+
+Continue Task A first.