From 136ff7400a6367265cd67b885809ff029deb02c1 Mon Sep 17 00:00:00 2001 From: "Alice (OpenClaw)" Date: Wed, 13 May 2026 18:43:28 +0800 Subject: [PATCH] =?UTF-8?q?docs:=20define=20remaining=20pluginization=20ta?= =?UTF-8?q?sks=20/=20=E5=AE=9A=E7=BE=A9=E5=89=A9=E9=A4=98=E6=8F=92?= =?UTF-8?q?=E4=BB=B6=E5=8C=96=20task=20=E8=88=87=E4=B8=8B=E4=B8=80?= =?UTF-8?q?=E5=80=8B=20migration=20=E7=9B=AE=E6=A8=99?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- reports/next-inline-migration-target.md | 36 ++++++++++++++++++++ reports/remaining-task-list.md | 45 +++++++++++++++++++++++++ 2 files changed, 81 insertions(+) create mode 100644 reports/next-inline-migration-target.md create mode 100644 reports/remaining-task-list.md diff --git a/reports/next-inline-migration-target.md b/reports/next-inline-migration-target.md new file mode 100644 index 0000000..81321df --- /dev/null +++ b/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 diff --git a/reports/remaining-task-list.md b/reports/remaining-task-list.md new file mode 100644 index 0000000..3d372d8 --- /dev/null +++ b/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.