feat(feishu): add group thread isolation lists

[luoye-fe]

Summary

Add per-group controls for Feishu thread_isolation so deployments can opt specific regular groups in or out without changing private-chat behavior. Native Feishu topic messages are always kept in topic-thread reply mode, even when global thread_isolation is disabled.

Type of change

• Bug fix (non-breaking change that fixes an issue)
• New feature (non-breaking change that adds functionality)
• Breaking change (fix or feature that would cause existing behavior to change)
• Documentation only
• Internal refactor / chore (no user-visible change)

Testing

Automated tests added in this PR

• TestNewFeishu_ThreadIsolationGroupLists in platform/feishu/platform_test.go — verifies the new Feishu platform options are parsed into the platform.
• TestFeishu_MakeSessionKeyThreadIsolationGroupLists in platform/feishu/platform_test.go — verifies global-on blacklist, global-off whitelist, native topic, private chat, and shared-channel session-key behavior.
• TestBuildReplyMessageReqBody_SetsReplyInThreadFlag in platform/feishu/platform_test.go — extended to verify blacklist/whitelist/native-topic reply-thread decisions.
• TestMarkAndIsActiveThreadSession in platform/feishu/feishu_test.go — extended to verify active thread tracking honors blacklist/whitelist behavior.

Commands run locally:

GOTOOLCHAIN=auto /opt/homebrew/bin/go test ./platform/feishu -run 'TestNewFeishu_ThreadIsolationGroupLists|TestFeishu_MakeSessionKeyThreadIsolationGroupLists|TestBuildReplyMessageReqBody_SetsReplyInThreadFlag|TestMarkAndIsActiveThreadSession' -v
GOTOOLCHAIN=auto /opt/homebrew/bin/go test ./platform/feishu

Both commands passed.

GOTOOLCHAIN=auto /opt/homebrew/bin/go test ./... was also run. It failed on unrelated local/environment-sensitive tests outside this change:

• core: TestCUJ_A3_ImageReachesAgent and TestCUJ_A5_FileReachesAgent failed during TempDir RemoveAll cleanup: directory not empty.
• daemon: TestLaunchdStatusUsesUserDomainWhenGUIDomainUnavailable failed with Status().Running = false, want true.

All Feishu platform tests passed in that full-suite run.

For bug fixes only — regression test

• Regression test name: N/A — this is a new feature.
• Manual verification this test catches the regression:
  • Reverted the fix locally; the regression test failed as expected.

Critical User Journeys (CUJ) impact

• No CUJ touched (small refactor, doc change, etc.)
• A — basic conversation
• B — session lifecycle (/new /switch /list /history etc.)
• C — agent execution control (/mode /cancel /stop permissions)
• D — security & permissions (allow_from admin_from banned_words rate limits)
• E — scheduled tasks (/cron /timer)
• F — config switching (/lang /provider /model reload)
• G — error handling & robustness (LLM failure, ws reconnect, agent crash)
• H — multi-platform / multi-project isolation
• I — UI rendering correctness (cards, streaming, display modes)

If any CUJ group is touched, confirm:

• go test ./core/ -run TestCUJ passes locally.
• If the change alters an existing user-visible flow, the corresponding
  CUJ test was updated (or a new CUJ added) to cover the new behavior.

Manual / user-visible behavior change

Feishu platform users can now configure:

• thread_isolation_black_group: when thread_isolation = true, listed group chat IDs keep normal non-topic replies.
• thread_isolation_white_group: when thread_isolation = false, listed group chat IDs use topic-thread replies.

Native Feishu topic messages continue to reply in the topic thread regardless of the global thread_isolation setting. Private chat behavior is unchanged.

Checklist (reviewer will verify)

• go build ./... passes
• go test ./... passes (with -race if touching concurrency)
• AGENTS.md Pre-Commit Checklist items are satisfied where applicable to this scoped Feishu platform change
• No new hardcoded platform/agent names in core/
• i18n strings have all-language translations (if any new user-facing text)
• No secrets / credentials in source

Related

• Issue:
• Related PR:

[luoye-fe]

@chenhg5 hi，帮忙 review 下这个 pr 吧，主要是有些群聊希望按话题形式回复，有些群聊希望用同一个上下文持续对话

[chenhg5]

结论: Approve

总体判断: 这是一个 well-scoped 的 Feishu platform 扩展, 给现有 thread_isolation 全局开关加上 per-group 黑白名单覆盖, 同时保证 native Feishu topic 消息始终走 topic-thread reply (跟现有 thread_isolation 设置解耦)。 实现方式正确 (helper 函数抽取 + 双重 session tracking + 完整 test 覆盖), 不破坏 backward compat, 4/4 CI green + 完整 PR 模板。 当前可合并。

Review 范围:

• 看了 4 files / +284/-12: platform/feishu/feishu.go (+245/-11) + config.example.toml (+4/-0) + 2 个 test 文件 (+35/-1)。
• 重点关注 correctness (per-group 覆盖逻辑) / native topic thread 解耦 / backward compat / test 覆盖 / 配置文件说明。

✅ 做得好的地方:

• 行为变更清晰: PR 描述明确给出 3 个新行为 — thread_isolation_black_group (global-on 时 excluded 群保持普通 reply) / thread_isolation_white_group (global-off 时 included 群走 thread reply) / native topic 始终走 thread (跟 global 设置解耦)。 每个行为都有对应 test。
• Helper 抽取合理 (platform/feishu/feishu.go:3346-3408): 5 个新 helper (shouldUseThreadIsolationForMessage / isNativeThreadMessage / shouldUseThreadIsolationForChat / configuredListContains / shouldTreatSessionAsThread) 每个职责单一, 跟核心 platform 文件的现有 helper 风格一致 (e.g. isAttachmentMsgType / stringValue)。 shouldUseThreadIsolationForMessage 把「native topic 检测」和「chat-level 配置」封装到一个函数, 主流程 makeSessionKey 调用变干净。
• Native topic 解耦做得对 (platform/feishu/feishu.go:174 + 3342-3352): 引入 nativeThreadSessions sync.Map 单独追踪 native topic 消息, 跟 activeThreadSessions 分开, 避免跟 global thread_isolation 状态混淆。 markNativeThreadSession / isNativeThreadSession 配对使用, 跟现有 markThreadSessionActive / isActiveThreadSession 模式一致。
• shouldUseThreadContext 集成自然 (platform/feishu/feishu.go:1438): 把 p.threadIsolation && isThreadSessionKey(sessionKey) 替换为 p.shouldUseThreadContext(sessionKey), 考虑了 native topic + chat 黑白名单两种情况, 主流程 dispatchMessage 改动最小 (只动一行)。
• Backward compat 完整保留: markThreadSessionActive 保留为 wrapper 调用新 markThreadSessionActiveForContext(sessionKey, false), 现有调用方不需要改; isActiveThreadSession / makeSessionKey 内部逻辑替换, 但外部 contract 不变 (group + global-on 仍按 thread 隔离, group + global-off 仍按 chat-id 共享 session)。
• Test 覆盖全面 (platform/feishu/platform_test.go + feishu_test.go):
  • TestNewFeishu_ThreadIsolationGroupLists — 验证新 option parsing
  • TestFeishu_MakeSessionKeyThreadIsolationGroupLists — 5 个 case (global-on blacklist / global-off whitelist / native topic / private chat / shared-channel)
  • TestBuildReplyMessageReqBody_SetsReplyInThreadFlag — extended 验证 blacklist/whitelist/native-topic 的 reply-thread 决策
  • TestMarkAndIsActiveThreadSession — extended 验证 active thread tracking 跟新规则一致
    每个 test 都有清晰的 case 命名, 跟前一轮 review 反馈一致 (test 应覆盖「global-on blacklist」+「global-off whitelist」两个对称 case)。
• PR 模板完整: 包括 Summary / Type of change / Testing (4 个 test + 命令) / CUJ impact (No CUJ touched) / Manual behavior change / Checklist / Related。 报告诚实标明「go test ./... 在 pre-existing 环境敏感 test 失败 (TempDir RemoveAll cleanup / launchd status), 跟本 PR 改动无关」。
• config.example.toml 注释完整 (config.example.toml:1008-1011): 每个新 option 都有中英双语说明 + 行为解释, 跟现有 Feishu config block 风格一致。

🔴 必须处理: 无。
🟠 建议改进:

• 🟠 P2: configuredListContains (platform/feishu/feishu.go:3389) 跟现有 core.AllowList 都有 trim/empty 处理逻辑, 看起来 configuredListContains 等价于 strings.TrimSpace(list) != "" && core.AllowList(list, value)。 建议直接用 core.AllowList 一个函数 (假定 core.AllowList 已经处理了 trim 和 empty), 减少一层间接。 但这不阻塞本 PR, 未来 cleanup 即可。
  验证: 检查 core.AllowList 源码确认是否已处理 trim/empty。
  影响: 微, 主要是函数调用层次清晰度。

🔵 可选优化:

• 未来考虑: 把 threadIsolationBlackGroup / threadIsolationWhiteGroup 拆成独立的 core.AllowListConfig struct 复用 (类似 allow_from), 跟 core.AllowList 配套。 但本 PR 用字符串 + 逗号分隔跟现有 allow_from 一致, 改动范围保持小是正确的取舍。

❓ 需要确认: 无。

Testing / Risk:

• 已验证: 5/5 CI checks PASS (lint 1m58s / unit-test 16m49s / smoke 33s / regression 31s / performance 48s) + 作者本地 go test ./platform/feishu + go test ./... (Feishu 部分全 PASS, 标出 pre-existing 环境失败)。
• 覆盖: option parsing + 5 case session-key (global-on blacklist / global-off whitelist / native topic / private / shared) + reply-thread decision + active thread tracking。
• 未覆盖 (不在本 PR 范围, 未来 follow-up): 黑白名单同时配置时的优先级 (没在 PR 描述里说, 看起来是「black > white, native 优先」但 code 没显式处理互斥配置)。
• 剩余风险: 低。 修改面限制在 Feishu platform 包内, 不动 core/engine/session。 backward compat 完整保留。

Next step:

• 作者: P2 是 nit, 跟 cleanup 一起做即可。 考虑未来 cleanup configuredListContains 直接用 core.AllowList。
• owner: 这是一个 additive feature, 不破坏现有部署, 可以安全 merge 进 v1.4.x。 建议 release notes 加一条 「Feishu: per-group thread_isolation 黑白名单 + native topic 始终 thread」, 方便运营根据具体群配置 reply 行为。
• QA: 本次 review 后会 pr set --workflow-state ready-to-merge + render。