ZLAgent

Fetch and summarise recent papers from arxiv.org. Use this skill when the user wants a digest of today's or this week's cs.AI / cs.LG / cs.CL listings, a focused per-topic roundup, or a quick look at a specific submission.

Lightweight code-review checklist the agent runs before declaring a patch ready. Catches the obvious classes of issue (dead code, missing error handling, weakened tests) without a full security audit. Adapted from addyosmani's "Review" lifecycle phase.

Root-cause-first debugging loop for code or runtime errors: understand the failure before proposing a fix. Adapted from Hermes' systematic- debugging — same iron law, smaller surface, ZLAgent-tool-aware.

Do not declare a task done without running verification. For code, that means running tests / smoke; for answers, that means checking the claim against a fresh source. Adapted from obra/superpowers and addyosmani's "Verify" lifecycle phase.

Umbrella skill for user document workflows — PDF, Word, Excel, PowerPoint, text extraction, summarization. Dispatches to the right MCP / tool; proposes an install via mcp-discovery when no local capability fits. Adapted from Anthropic's docx/pdf/xlsx/pptx skills.

When the user wants something the agent doesn't have a tool for yet (travel planning, GitHub issues, Linear, weather, filesystem, ...), search the live MCP ecosystem for an existing server, propose installing it, and on confirm wire it up via mcp_manage. Adapted from Hermes' mcporter skill — same flow, ZLAgent-tool-aware.

定期检查 ZLAgent 的 skill、MCP 工具、记忆摘要的变化， 生成一份变更摘要并通过 IM 发送给用户。

Produce concise docs and Architectural Decision Records (ADRs) when a workflow or design becomes durable. Favours short artifacts over exhaustive manuals. Adapted from addyosmani's "documentation-and-adrs" and the skill-authoring doctrine.

Author a new ZLAgent skill in the canonical Trigger / Inputs / Steps / Verification / Failure Signals format. Class-level only. Used when the user says "把这个流程记成 skill" or in background skill review.

在安装来自外部 URL / GitHub / 他人分享的 SKILL.md 之前， 对其内容做安全性和质量审查，输出审查报告和安装建议。

Meta skill that guides the agent on when to pick an existing skill, when to create a new one, and when to install an MCP server. Adapted from obra/superpowers' "using-superpowers" and addyosmani's "meta" skill for personal-AI use.

Find a small list of academic papers on demand — "给我 N 篇目标检测领域 顶刊 / 顶会 / SOTA 论文 / 综述" — using the multi-source chain arxiv API → Semantic Scholar → web_search. Distinct from `daily-paper-digest` (scheduled daily delivery) and `arxiv` (single-day category dump): this one is the on-demand "find me N papers about X" surface that the user usually triggers in chat.

Reply with a structured plan instead of executing. Goal, assumptions, numbered steps, verification, risks. No file writes unless the user explicitly asks. Used when the user says "先列个计划" / "先想想再做".

Scheduled or on-demand digest of recent papers in a chosen field, with source URLs and a one-line takeaway per paper. Used as the skill_hint for cron jobs like "每天 8 点给我发最新论文".

Compact news roundup on a topic: web_search for fresh items, dedup, read 1-2 sources per item, return short bullets with links. Use as the skill_hint for "每天 8 点给我发新闻" cron jobs.

Look up current weather, short forecast, or rain check for a city using wttr.in. No API key, no new tool — uses read_url against https://wttr.in/<city>?format=...

Summarise a single web URL the user shares: fetch with read_url, produce a compact key-points reply. No external CLI, no API key.

Generate a single rich-card travel itinerary on demand. The user names a destination plus a duration ("<DEST>三日游", "<DEST>赏樱攻略", "<DEST>美食推荐") and the agent returns ONE structured RichMessage covering route, food, transport, stay, budget, pitfalls, and packing. v0.38 replaced the v0.37.5 multi-section streaming flow with a single rich-content card, which (a) avoids WeChat rate limits (used to fan out 6+ messages), (b) renders as a clean markdown card on wecom_bot, (c) renders as CJK-aligned plain text on weixin (no more broken markdown tables).

Handle time-sensitive travel requests by using built-in realtime travel lookup or attaching MCP servers for railway tickets, flights, maps, local places, weather, hotels, and route planning. Prefer a quick answer path: call the built-in travel_realtime tool first for 12306/Open-Meteo; use existing mcp__* tools next; otherwise propose one safe candidate and install only after confirmation.

对 AI 生成的文本做去 AI 味润色：去掉套话、模板感、 过度完整的列表结构，让输出更像人话、更自然。