Qwen Code v0.19.5

<!-- Release notes generated using configuration in .github/release.yml at release/v0.19.5 --> ## What's Changed * feat(cli): Harden daemon-managed channel worker by @doudouOUC in https://github.com/QwenLM/qwen-code/pull/6098 * fix(web-shell): defer session creation until first prompt by @ytahdn in https://github.com/QwenLM/qwen-code/pull/6066 * docs: document model/auth settings, /model --vision, and --safe-mode by @DragonnZhang in https://github.com/QwenLM/qwen-code/pull/6028 * fix(cli): clip live markdown to the viewport to stop non-VP scrollback replay by @chiga0 in https://github.com/QwenLM/qwen-code/pull/6081 * ci: add fork PR safety precheck by @yiliang114 in https://github.com/QwenLM/qwen-code/pull/5926 * feat(serve): support HTTPS/TLS via --tls-cert and --tls-key flags by @pomelo-nwu in https://github.com/QwenLM/qwen-code/pull/6032 * fix(cli): yield to React after addItem to reduce input lag by @Alex-ai-future in https://github.com/QwenLM/qwen-code/pull/6059 * ci: persist npm cache on self-hosted runners by @yiliang114 in https://github.com/QwenLM/qwen-code/pull/6130 * feat(cli): show description and level in /skills ACP output by @doudouOUC in https://github.com/QwenLM/qwen-code/pull/6117 * fix(diff): show whitespace-only edits instead of 'No changes detected' by @doudouOUC in https://github.com/QwenLM/qwen-code/pull/6141 * Add compact session timeline rail by @callmeYe in https://github.com/QwenLM/qwen-code/pull/6078 * feat(core,cli): unified reasoning effort with /effort command by @DragonnZhang in https://github.com/QwenLM/qwen-code/pull/6072 * feat(core): Add leader approval for plan-required teammates by @doudouOUC in https://github.com/QwenLM/qwen-code/pull/6138 * fix(ci): grant PR review precheck permissions by @wenshao in https://github.com/QwenLM/qwen-code/pull/6147 * fix(ci): list workflow comments with GET by @wenshao in https://github.com/QwenLM/qwen-code/pull/6148 * fix(ci): use CI_BOT_PAT for precheck comment on fork PRs by @yiliang114 in https://github.com/QwenLM/qwen-code/pull/6151 * fix(cli): Avoid blocking WebUI sessions on MCP readiness by @BZ-D in https://github.com/QwenLM/qwen-code/pull/6161 * fix(ci): allow prechecked fork PR automation by @yiliang114 in https://github.com/QwenLM/qwen-code/pull/6160 * fix(ci): fall back to latest autofix sandbox image by @yiliang114 in https://github.com/QwenLM/qwen-code/pull/6159 * fix(ci): create precheck comments via REST by @wenshao in https://github.com/QwenLM/qwen-code/pull/6156 * fix(web-shell): show skill slash commands (e.g. /review) before first prompt by @wenshao in https://github.com/QwenLM/qwen-code/pull/6153 * docs: document the /config slash command by @DragonnZhang in https://github.com/QwenLM/qwen-code/pull/6145 * fix(web-shell): only show scroll-to-bottom button when content overflows by @ytahdn in https://github.com/QwenLM/qwen-code/pull/6150 * fix(web-shell): improve disconnected composer handling by @ytahdn in https://github.com/QwenLM/qwen-code/pull/6166 * fix(ci): diagnose autofix publish credentials by @yiliang114 in https://github.com/QwenLM/qwen-code/pull/6162 * fix(release): reduce npm package scan triggers by @yiliang114 in https://github.com/QwenLM/qwen-code/pull/6164 * fix(channels): replace setTimeout(0) drain with turn_complete SSE barrier by @doudouOUC in https://github.com/QwenLM/qwen-code/pull/6165 * feat(channels): add DingTalk proactive send for channel loops by @qqqys in https://github.com/QwenLM/qwen-code/pull/6174 * fix: lazy-load memory prompt when indexes are empty (#6097) by @kagura-agent in https://github.com/QwenLM/qwen-code/pull/6104 * feat(channels): add identity and task lifecycle metadata by @qqqys in https://github.com/QwenLM/qwen-code/pull/6105 * feat(core): add retry with backoff for MCP capability discovery by @DennisYu07 in https://github.com/QwenLM/qwen-code/pull/6158 * fix(cli): skip MCP approval dialogs in YOLO mode by @qwen-code-dev-bot in https://github.com/QwenLM/qwen-code/pull/6177 * feat(channels): add listSessions to ChannelAgentBridge by @doudouOUC in https://github.com/QwenLM/qwen-code/pull/6182 * ci: Add prepare-pr skill for autofix PR bodies by @yiliang114 in https://github.com/QwenLM/qwen-code/pull/6184 * feat(channels): show lifecycle status in adapters by @qqqys in https://github.com/QwenLM/qwen-code/pull/6114 * fix(serve): keep skill slash commands available when the ACP child is unavailable by @wenshao in https://github.com/QwenLM/qwen-code/pull/6169 * fix(web-shell): polish session timeline rail by @callmeYe in https://github.com/QwenLM/qwen-code/pull/6171 * test: stabilize plan mode tool-control E2E by @yiliang114 in https://github.com/QwenLM/qwen-code/pull/6176 * fix(web-shell): mobile UX — safe areas, overscroll, native-app feel by @pomelo-nwu in https://github.com/QwenLM/qwen-code/pull/6142 * feat(web-shell): overhaul list-dialog interaction, keyboard nav & a11y by @carffuca in https://github.com/QwenLM/qwen-code/pull/6128 * feat(cli): add credential redaction for worker stderr forwarding by @doudouOUC in https://github.com/QwenLM/qwen-code/pull/6146 * fix(cli): drop /effort tier autocompletion for an argument-hint placeholder by @DragonnZhang in https://github.com/QwenLM/qwen-code/pull/6179 * fix(ci): limit fork PR precheck to safety signals by @yiliang114 in https://github.com/QwenLM/qwen-code/pull/6178 **Full Changelog**: https://github.com/QwenLM/qwen-code/compare/v0.19.4...v0.19.5