Explore skills
Search every indexed skill by name, description, repository or instructions, then narrow the results.
Showing 32131–32160 of 34699 skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| aikey-reframeiamzifei/aikeyAI KEY·越改越不对 ——改了好几版还是不对的时候用。不急着出下一版,先跳出来看:这个问题本身问得对不对、手上哪些做法其实可以不做、毛病到底出在哪一层,再在同一条回复里直接给能用的改法。稿子、标题、选题、技能和提示词、生意上的事都适用。只动定位到的那一处,你原来的判断和想要的结果不替你换。 触发方式:/aikey-reframe、/升维、/跳出来看、「升维」「站高一点看」「是不是只在局部打转」「改了五六版还是不对」「越改越差」「补丁越打越多」「问了几个 AI 各说各的」「两个要求打架」 For when you have revised the same thing several times and it is still wrong. Steps back first: is the question itself right, which planned actions can be dropped, which layer the problem actually sits in — then gives the fix in the same reply, touching only the located spot. Trigger: /aikey-reframe, "step back", "is this a local problem", "still wrong after five revisions" —— AI KEY · 不给公式,给判据。每条规则都标了实测代价。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| aikey-resonateiamzifei/aikeyAI KEY·戳不戳得中人 ——共鸣诊断技能。不从理论出发,从观众出发:这条内容让观众带走了哪一种收获(原来是这样 / 我就说吧 / 他替我说了),会不会动手(收藏 / 转发 / 评论 / 关注),每个判断都指到原文并标它是证据、推断还是待验证。模式A 诊断自己的草稿;模式B 拆一条别人的爆款反推可借的角度;模式C 判一句「我觉得不对劲」背后是一件事还是一群人的事。与 aikey-review 正交:review 防句废,resonate 防结构散。不负责逐句红线和能不能发(→ aikey-review)、哪一拍会划走(→ aikey-flow)、写标题或开头(→ aikey-title / aikey-hook)。 触发方式:/aikey-resonate、/戳不戳人、/共鸣、/aikey-共鸣、「这稿有没有戳中人」「会不会没人看」「这条为什么能火」「拆一下这个爆款」「受众到底想听什么」「这事值不值得拍一条」「我总觉得哪里不对但说不上来」「这是我一个人的事还是大家的事」 Resonance diagnosis, audience-first: which takeaway the piece delivers, which action impulse it triggers, with every claim tied to a quoted line and labelled evidence / inference / to-verify. Mode A: your draft. Mode B: decode a hit. Mode C: test whether a hunch is one incident or a pattern. Not for: line-level red lines and publish-readiness (aikey-review), drop-off beats (aikey-flow), writing titles or openings (aikey-title / aikey-hook). Trigger: /aikey-resonate, "will this resonate", "why did this blow up", "decode this viral post" —— AI KEY · 不给公式,给判据。每条规则都标了实测代价。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| aikey-retroiamzifei/aikeyAI KEY·发布后复盘 ——短视频复盘技能。发布后把真实平台数据收回来,对照发布前的预判做归因讨论,把验证过的规律写进技能记忆——让系统学会「为什么这条火」。也可以一次回看最近一批已发内容,看它们合起来给人留下了什么印象、和你想让人记住的差多远。 触发方式:/aikey-retro、/数据出来了、/复盘、/aikey-复盘、「这条数据出来了」「为什么这条火」「为什么这条没火」「复盘一下」「最近这些条给人什么感觉」「账号是不是跑偏了」 Post-publish retro: collect real platform data, attribute against pre-publish predictions, write validated patterns into skill memory. Also reviews a recent batch to see what impression the posts add up to. Trigger: /aikey-retro, "the numbers are in", "why did this video flop/pop" —— AI KEY · 不给公式,给判据。每条规则都标了实测代价。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| aikey-revenueiamzifei/aikeyAI KEY·这个月钱去哪了 ——营收异动归因。这个月钱少了(或多了),到底是客人变少、每人买得少、还是单价变了——三种原因的处理动作完全相反,分不清就会做反。也识别「静默侵蚀」:每个月只跌一点点、单月都像正常波动,累计起来致命。 触发方式:/aikey-revenue、/钱去哪了、/营收归因、「这个月为什么少了」「营收掉了」「钱去哪了」「生意怎么突然不行了」「涨了但我不知道为什么」 Revenue movement attribution for owner-operators. Splits any change into customer count × per-customer volume × price, so the right fix follows. Also detects slow erosion that hides inside monthly noise. Works from a ledger or order book — no dashboard required. Trigger: /aikey-revenue, "why is revenue down this month", "where did the money go", "business suddenly slowed" —— AI KEY · 不给公式,给判据。每条规则都标了实测代价。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| aikey-reviewiamzifei/aikeyAI KEY·发布前审一遍 ——口播稿发布前审核技能。按观众的四次决定审:点不点进来 · 留不留下来 · 记不记得你 · 做不做点什么。先盲读开头判留走,逐句信息密度评分(60/80 分线)+ 十一问 + 六种反应预测(只算观众走得到的部分,含流量是谁、离生意多远)+ 红线六查(改法给稳妥版和保留力度版两版)+ 机器信号层(导流 / 广告形状 / 名单词,与内容违规分开报),默认只诊断不改。不负责重写整稿(→ aikey-script)、换开头或标题(→ aikey-hook / aikey-title)、拆别人的爆款(→ aikey-resonate)。 触发方式:/aikey-review、/能不能发、/审核、/aikey-审核、「这稿子能不能发」「帮我审一下」「过一遍红线」「信息密度够不够」 Pre-publish review for talking-head scripts, organised around the viewer's four decisions: click, stay, remember, act. Per-sentence density scoring, eleven questions, red-line audit. Diagnose-only by default. Not for: rewriting the script (aikey-script), replacing the opening or title (aikey-hook / aikey-title), decoding someone else's hit (aikey-resonate). Trigger: /aikey-review, "can I publish this", "review my script" —— AI KEY · 不给公式,给判据。每条规则都标了实测代价。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| aikey-scriptiamzifei/aikeyAI KEY·口播稿写作 ——口播稿协作写作技能(知识付费口播为主形态)。三类脚本选型 + 从内容单元装配 + 逐段共创,不一键成稿。 触发方式:/aikey-script、/写口播、/写稿、/aikey-写稿、「写个口播稿」「帮我把这个题写出来」「知识付费口播」「写条视频文案」 Collaborative koubo (talking-head) script writing: script-type selection, unit assembly, section-by-section co-writing. Trigger: /aikey-script, "write a talking-head script", "turn this topic into a video script" —— AI KEY · 不给公式,给判据。每条规则都标了实测代价。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| aikey-skillifyiamzifei/aikeyAI KEY·做成一个技能 ——把这次会话里已经跑通的做法固化成一个新技能。不是从想法造技能——是从**已经产生过正确结果的那一段过程**里提炼,所以只在事情做完之后用。 触发方式:/aikey-skillify、/固化、/做成技能、/aikey-固化、「这次的做法留下来」「把刚才那套变成技能」「下次别再重新想一遍」「这个流程以后还要用」 Turn a method that already worked in this session into a reusable skill. Extracts from a completed run, never from an idea — so it only fires after the work is done. Trigger: /aikey-skillify, "make this a skill", "save this workflow", "I don't want to re-derive this next time" —— AI KEY · 不给公式,给判据。每条规则都标了实测代价。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| aikey-titleiamzifei/aikeyAI KEY·标题与封面 ——多平台标题与封面。没指定平台时默认按抖音 → 视频号 → 小红书出齐,一屏交付:每个平台首选标题 + 2 个备选、一版封面字,抖音另给一版作品描述,各附一句大白话理由;说「展开」再给全部候选和分析。按内容本身的形状选结构,过卖真 / 违禁词 / 不点名 / 黑话红线。**不给「通用爆款公式」:标题结构在不同账号间不通用,附脚本让你算出自己账号的规律。** 先看片子拍到哪一步,拍完的只承诺片子里有的。不负责口播开头(→ aikey-hook)、封面图片(→ aikey-cover)、能不能发(→ aikey-review)。 触发方式:/aikey-title、/封面、/标题、/aikey-标题、「起个标题」「抖音封面写什么」「小红书标题」「封面大字怎么写」「帮我优化这个标题」 Multi-platform titles and cover text in one screen: a pick plus two alternatives per platform, one cover, one caption; full candidates on request. Picks a structure that fits the content rather than a "proven formula". Not for: opening lines (aikey-hook), cover images (aikey-cover), publish-readiness (aikey-review). Trigger: /aikey-title, "give me a title", "Douyin cover text", "xiaohongshu title" —— AI KEY · 不给公式,给判据。每条规则都标了实测代价。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| aikey-topiciamzifei/aikeyAI KEY·今天拍什么 ——短视频选题技能。先探寻现象、检验假设、比较解释,把问题转成观众在意的选择。交互式选题:供需判定 + 对标信号 + 本人真实数据佐证 + 配比检查,不一键生成选题清单。 触发方式:/aikey-topic、/拍什么、/选题、/aikey-选题、「今天拍什么」「帮我出选题」「这个题能不能做」「有个想法你帮我判断下」「这题只有一头」「这题没劲」「帮我把它讲出两头来」 Topic selection for short videos: supply-demand test, benchmark signals, real data evidence. Trigger: /aikey-topic, "what topic should I shoot", "is this topic worth making" —— AI KEY · 不给公式,给判据。每条规则都标了实测代价。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| aikey-trackiamzifei/aikeyAI KEY·有什么到期了 ——在途与到期。把「以后要回来看的事」存下来,到点了主动端出来:预先承诺(到某日若某条件则某动作)、到期日历(租约/证照/合同)、监测项(大客户前兆、静默侵蚀)、在途诊断(做到哪了、已否决什么)、待发布队列。到期时只问条件满没满足,不重开辩论。 触发方式:/aikey-track、/有什么到期了、/盯、「记下来到时候提醒我」「有什么到期了」「上次做到哪了」「这个结论存一下」「该复查了吗」 Cross-session tracking for commitments, expiry dates, watch-items, in-flight diagnoses, and publish queues. Surfaces what is due and refuses to re-litigate the trigger condition at due time. Trigger: /aikey-track, "remind me to check this later", "what's due", "where did we leave off", "save this conclusion" —— AI KEY · 不给公式,给判据。每条规则都标了实测代价。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| aikey-trendiamzifei/aikeyAI KEY·风口在哪 ——一个大变化正在发生(AI 来了、平台换了、行业在洗牌),想知道它到底会走到哪、自己该押什么的时候用。不给「风口清单」,也不陪你预言未来 —— 带你做八个动作:从一场已经发生的颠覆倒推机制 → 把趋势拆成几条能量的曲线 → 看拆完之后还剩哪一两项没解决 → 找一个已经走完的老产业当剧本 → 先让自己活在未来里一段时间 → 落成「赌什么、不赌什么」→ 分开「趋势对不对」和「你能不能赢」→ 每个判断挂一条到期会自己回来检验的信号。 触发方式:/aikey-trend、/风口在哪、/看大势、「AI 来了我该押什么」「这是不是风口」「我这行会被 AI 怎么改」「下一个机会在哪」「现在入场晚不晚」「这波趋势能持续多久」 For when a big shift is under way and you want to know where it lands and what to bet on. No list of hot sectors and no fortune-telling: work back from a disruption that already happened, break the trend into curves you can measure, find what is still unsolved, borrow an industry that already ran the same script, live in the future for a while, turn the view into bets and non-bets, separate "is the trend right" from "can I win", and attach a dated signal to every call. Trigger: /aikey-trend, "AI is here, what should I bet on", "is this a real wave", "how will AI change my industry", "am I too late" —— AI KEY · 不给公式,给判据。每条规则都标了实测代价。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| keyiamzifei/aikeyAI KEY·短命令 ——AI KEY(AI 钥匙)工具箱的短命令。`/key` 和 `/aikey` 完全一样,只是少敲三个字母 —— 它把你交给总入口,由总入口按你卡在哪挑技能。 两套技能:做内容(选题 / 写稿 / 标题 / 开头 / 审核 / 剪辑 / 复盘)· 看生意(营收归因 / 大客户集中度 / 依赖风险 / 多产品线取舍 / 目标拆路径 / 趋势押什么 / 决策 / 到期追踪)。 触发方式:/key、/key 新手指南、/钥匙 ⚠️ **与 API key、密钥、密码、令牌无关**。用户问的是接口密钥、环境变量、凭据一类的事,**不要用本技能**。 Short command for the AI KEY toolbox: `/key` is the same entry point as `/aikey`. Not about API keys, secrets or credentials. Trigger: /key, "/key 新手指南" —— AI KEY · 不给公式,给判据。每条规则都标了实测代价。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| academic-paper-revisionZoeLinUTS/Academic-paper-revisionRevise an academic research paper draft (LaTeX) toward a specific target venue's conventions and a fluent, accepted-paper writing style. Use this whenever the user wants to prepare, tighten, restructure, or "clean up" a paper for submission — e.g. "get this ready for EMNLP", "make this read like an accepted ACL paper", "trim to the page limit", "remove the RQ/RO scaffolding", "add an algorithm/equation for the method", "why does this look like a student draft", "evaluate the novelty", "how did this line of work develop / what's the prior work / how is this different", or when they mention a venue (ACL, NAACL, EMNLP, AJCAI, NeurIPS, etc.) together with a manuscript. Trigger even if the user does not say the word "revise" — any request to make a paper draft match a venue or read more like a real published paper should use this skill. Accepts a target venue as input. | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| hiapi-seedance-2-5-videoHiAPIAI/hiapi-seedance-2-5-video-skillPlan, generate, recover, download, and quality-check Seedance 2.5 videos through HiAPI. Use for Seedance 2.5 text-to-video, first/last-frame image-to-video, multimodal reference-to-video, or a complete brief-to-delivery video workflow. | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| biologyScilifelabDataCentre/scilifelab-ai-agent-mcp-workshop-2026-11-02Proteins, drug targets and peptides. Look up protein function, location and diseases (UniProt), a drug's targets, mechanism of action and indications (ChEMBL), protein interaction partners (STRING), fetch protein sequences, compute sequence properties and align sequences (Biopython). Use for any question about targets, mechanisms, protein biology or protein/peptide sequences. | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| chemistryScilifelabDataCentre/scilifelab-ai-agent-mcp-workshop-2026-11-02Small-molecule cheminformatics with RDKit. Resolve compound names or IDs to SMILES, remove salts, compute physicochemical properties, check drug-likeness rules (Lipinski, Veber), compare structural similarity (Tanimoto) and search for substructures. Use for any question about a compound's structure, properties, drug-likeness or analogues. | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| literatureScilifelabDataCentre/scilifelab-ai-agent-mcp-workshop-2026-11-02Search PubMed for evidence with the NCBI LitSense2 semantic-search API. Returns passages from papers with their PMIDs. Use whenever an answer needs published evidence or citations, e.g. mechanism of action, resistance, clinical findings, or PK parameters to feed into a simulation. | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| pkpdScilifelabDataCentre/scilifelab-ai-agent-mcp-workshop-2026-11-02Pharmacokinetic / pharmacodynamic simulation with SciPy. Concentration-time profiles for repeated oral or IV dosing (Cmax, trough, AUC, half-life, steady state), Emax effect models, dose finding, regimen comparison, loading or missed doses and custom compartment models. Use for any dosing, exposure or target-coverage question. | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| adversarial-gateMarchTalk/ai-content-skills独立上下文的怀疑式审核,两种模式,串行调用(不要并行起两个 Agent,会撞速率限制): · mode=judgment-stress(命题定稿后、写正文前)——起一个全新上下文的 agent, 对刚定稿的判断做压力对抗:逻辑链有没有断裂、有没有明显反例、是不是只停在表面没挖到机制/人性层、 是不是新瓶装旧酒。测的是论证站不站得住,不是立场讨不讨喜——反共识判断不因为"反共识"被扣分。 · mode=post-draft-audit(终稿、quality-gate 通过后)——独立上下文核验终稿有没有真的把判断往深处 逼了还是只是包装了已有想法(内容层面的谄媚检测),并核对 judgment-stress 阶段提出的质疑有没有 被真正解决,同时可选做结构查重(比对你自己的历史稿件记录)。 两种模式默认姿态:先假定平庸,除非稿子自己证明不是。禁止清单打勾式审查——每次质疑必须从这条内容本身 现场生成,不能套固定题库。 触发:/adversarial-gate、「压力对抗这个判断」「这条是不是谄媚我」「独立审核一下」、 写稿流程里命题定稿后/终稿质检后自动调用 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| buyer-psychologyMarchTalk/ai-content-skills商业心理与购买阻力底层能力(合并自两个原本总是配对使用的判断层:经济理性层锁定阻力分类,心理感受层翻译成内心独白)。先用六类阻力锁定用户卡在哪(钱/风险/时间/信任/面子/决策成本),再翻译成用户不会说出口的心理独白和内容锚点。处理产品价值、价格、信任、成交、风险、成本、用户为什么不买;区分伪需求和真实付费需求、流量内容和成交内容;落点必须是钱、风险、时间、信任、结果、代价,以及具体的一句话/一个场景/一个动作。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| concept-checkMarchTalk/ai-content-skills概念核验。两种深度:A 贴身词语核验(写文案时一个词该不该留,三问自检 + 行业黑话豁免);B 深度概念拆解(把一个模糊概念拆到说不清就是没想清——对齐是否各说各话、往回倒查原意、假设检查戳穿伪概念、检验是描述行动还是替代了行动、去词测试、判断是信息缺口还是执行缺口)。别只停在"是不是 AI 空词"就收工,那只是 A 模式;用户要"深度拆解"必须走完 B 模式。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| content-positioningMarchTalk/ai-content-skills账号定位、人设定位、内容栏目、目标用户、承接路径。先锁定目标用户避免吸引错人,把产品优势翻译成用户场景,区分个人 IP / 创始人 IP / 品牌 IP / 门店 IP / 工厂 IP。内容支柱允许商业向(服务成交)和人设向(服务 IP 建立)并存。输出必须包含人设、目标用户、内容支柱、承接路径。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| material-researchMarchTalk/ai-content-skills实料调研·防口水话。任何写稿动笔前必跑的强制前置闸门。每次都跑,但按内容类型分流: · 干货 / 科普 / 数据 / 测评 / 盘点 / 教程 / 平台玩法型 → 强制 WebSearch / WebFetch 真查具体案例 / 数字 / 最新动态 / 一手细节,列「实料清单」,清单为空不许往下写,正文 ≥ 3 处可追溯硬料。 · 观点 / 认知 / 人设 / 价值观 / 思考型 → 判定靠第一手判断、无需外部二手料,显式放行。 额外一步:命题定稿后可选做「内核深化」——查这个判断相关领域最强的反方观点,不为引用,只为把判断本身逼深。 调研 ≠ 拼凑报告:判据是「删掉这条料,稿子会不会塌一块」,会才留。 触发:/material-research、写稿流程动笔前自动前置、「查料」「这条要不要查」「防口水话」「写得太泛」 Mandatory real-material research gate. Runs before every write; routes by content type—facts-heavy types must web-research and build an evidence list; opinion/persona types pass through on first-hand judgment. | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| organizeMarchTalk/ai-content-skills深挖对话的整理成稿工具。深挖已经完成(不是从零开始写)之后,把对话里的实质内容忠实、完整地整理成正式稿件——不做精选,不设默认字数,保留追问-反驳-修正的张力,张力归属必须忠于事实。 触发:/organize、「整理成稿」「把这个整理一下」「深挖完了帮我整理」「把这段深度拆解的内容整理」 Deep-dig-to-draft organizer. After a free-form dig is done, faithfully and completely organize the dialogue's substantive content into a formal draft — no curation, no default length, preserving argumentative tension with accurate attribution. | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| platform-distributionMarchTalk/ai-content-skills抖音、小红书、视频号的内容适配。抖音抓前 3 秒 / 冲突 / 节奏 / 完播 / 评论情绪;小红书抓封面点击 / 标题关键词 / 搜索词 / 前几行 / 收藏理由;视频号靠熟人信任 / 老板人设 / 长期口碑 / 朋友圈联动。同一选题必须能改成不同平台版本。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| quality-gateMarchTalk/ai-content-skills最终质检。按内容定位(流量 / 信任 / 成交 / 人设 / 思考)分别用对应的质检清单。检查空话、模板腔、AI 腔、重复句式、目标用户错位、是否缺场景 / 代价 / 信任锚 / 承接(成交型)或人设印象(人设型)。给修改意见,必要时直接重写。人设 / 价值观 / 思考型内容不挂商业承接不算缺陷。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| topic-strategyMarchTalk/ai-content-skills选题策略,决定什么值得写。从用户处境、行业矛盾、成交阻力、价格争议、客户误解、人设观察、价值观、长期思考找选题。每个选题必须标注生态资产、定位类型、目标用户、平台、流量理由 / 共鸣理由、承接或人设目的、风险点。承接按内容类型处理,不强制每条围绕变现;认知、观点、故事、人设、价值观、思考类内容可以不挂商业承接。 | Stars: 4 | Updated | Claude CodeCodex | Unverified | Install |
| comprehension-present-missing-understandingjacob-balslev/skill-graphNegative fixture for the gate-conformance suite: the audit-state sidecar declares `comprehension_state: present` but the SKILL.md omits the five flat Understanding fields. Activate only inside test-gate-conformance.js to prove the cross-file lint gate (comprehension_state -> Understanding fields) FAILS. Do NOT use as a production skill. | Stars: 3 | Updated | Claude CodeCodex | Unverified | Install |
| dangling-verify-withjacob-balslev/skill-graphNegative fixture for the gate-conformance suite: valid v8 frontmatter except `relations.verify_with` points at a skill that does not exist. Activate only inside test-gate-conformance.js to prove the structural lint gate FAILS on a dangling relation target. Do NOT use as a production skill. | Stars: 3 | Updated | Claude CodeCodex | Unverified | Install |
| invalid-sidecarjacob-balslev/skill-graphNegative fixture for the gate-conformance suite: the SKILL.md frontmatter is fully valid, but the sibling audit-state.json sidecar is missing the required `owner` field. Activate only inside test-gate-conformance.js to prove the structural lint gate validates the sidecar against schemas/skill-audit-state.schema.json. Do NOT use as a production skill. | Stars: 3 | Updated | Claude CodeCodex | Unverified | Install |