QQ delivers emoji messages as opaque <faceType=..,faceId="..",ext="..">
markup fragments (the ext field holds a base64 JSON blob with the face
name, e.g. {"text":"吃瓜"}). Without translation, the Harness receives
only this unusable tag and loses what the sender actually meant.
Register the SDK's contentSanitizer({ parseFaceTags: true }) middleware,
mirroring tencent-connect/dsh-qqbot, so C2C and group inbound messages
surface as e.g. 【表情: 吃瓜】 instead of the raw tag.
- qq-runtime.mjs: register contentSanitizer before the typing indicator
- runtime.test.mjs: assert sanitizer registration and update middleware order
- lib/index.js: regenerated host bundle
The Feishu bot now calls the app_slash_commands OpenAPI at startup to register its common commands (menu, new, help, status, compact, sessionlist, workspacelist, watch, unwatch, watchlist, archived) as native Slash Commands, so typing / in a Feishu direct-message input box pops the command panel and tapping a command triggers it.
The command list is owned and pushed by dsh-im; it does not depend on the dsh/Harness backend. Registration is best-effort and runs asynchronously, so it never blocks the long-connection startup or message delivery when the app lacks the application:app_slash_command:read/write scopes.
Adds src/channels/feishu/slash-command-registry.mjs (manifest, list/create/delete, idempotent sync), wires it into feishu-runtime.mjs (slashCommands config flag + status fields), and covers it with tests.
Reuse sender metadata and the existing AI Card lifecycle to include native mentions in ordinary group answers. Preserve mentions across progress, completion, and cleanup without sending a separate reminder.
Verified with npm run check (1671 tests) and a real ordinary Q&A in the existing DingTalk test group.
Add regression coverage for group mention chunks, private replies, mention payloads, and batch reply failure handling. Credit divingleee in package contributors and rebuild the checked-in bundles.
Keep a bounded streaming preview, then deliver the complete final answer across multiple cards. Preserve all provider message IDs and cover long snapshots and Unicode boundaries with regression tests.
Fixes#78
Add profile-scoped npm updates for Desktop and Web, protect source installations, and keep restarts manual. Cover stale status recovery, request races, and dialog focus with regression tests.
Verified with npm run check (1626 tests) and isolated Desktop/Web npm installation flows.
Refs #61
Add shared /history parsing, read-only session access, message filtering, and bounded replies. Default to three messages, cap at five, and exclude saved history commands.
Include channel integration tests, translations, command help, and the Issue #62 implementation plan.
Keep images for newer Harness endpoints and retry without it only when an older gateway rejects the field before dispatch. Preserve cancellation and never retry business failures.
Add regression coverage, rebuild the Host bundle, and document the compatibility fix. Follow-up to #75.
The Harness commands execute descriptor declares agentId, line, and images
as required JSON wire fields (strict codec, no acceptsUndefined). The Host
command executor only sent agentId and line, so the Typert gateway rejected
every IM /compact call with arguments-invalid before the command could run.
Send images: [] so /compact works from every IM channel.
Treat attachment-free Feishu rich-text posts containing bot commands like ordinary text commands instead of forwarding them to the Harness.
Co-authored-by: OpenClaw Agent <agent@openclaw.ai>
