[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M754433ZCS4YZDBWDWNM5K0N":3,"skill-same:01M754433ZCS4YZDBWDWNM5K0N":58},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":28,"commitSha":29,"skillDirectory":30,"forksCount":31,"repositoryHtmlUrl":32,"repositoryHomepageUrl":33,"ownerHtmlUrl":34,"ownerBlogUrl":35,"ownerDescription":36,"updatedAt":37,"install":38,"resources":51},"01M754433ZCS4YZDBWDWNM5K0N","data-digest",null,"Use when asked for a recurring data digest, a \"what changed since last time\" update, a weekly metrics readout, or a trend\u002Fanomaly check on a CSV, spreadsheet, SQL result, or connected data source — compare against the prior run, lead with the deltas, and call out anomalies.","matangilabs\u002Fkortny","matangilabs","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F293936513?v=4","01M75421VCEBXSQNHDWZSMQS1N","Matangi Labs","01M75421WCPAD845MDGEMK8KFM",6,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fmatangilabs\u002Fkortny\u002Fblob\u002Fmain\u002Fkortny\u002Fskills\u002Fcurated\u002Fdata-digest\u002FSKILL.md",[],[],"2026-11-12 13:55:43",false,"---\nname: data-digest\ndescription: Use when asked for a recurring data digest, a \"what changed since last time\" update, a weekly metrics readout, or a trend\u002Fanomaly check on a CSV, spreadsheet, SQL result, or connected data source — compare against the prior run, lead with the deltas, and call out anomalies.\nmetadata:\n  version: 1.0.0\n  display_name: Data Digest\n  tags: data, digest, recurring, metrics, trends, anomalies, deltas\n---\n\n## Goal\n\nProduce a recurring data readout whose value is *comparison*: what moved since last time, what's anomalous, what's worth acting on — across whatever source the data lives in.\n\nThis is the recurring, longitudinal cousin of `data-brief`. `data-brief` turns one posted file into a one-shot story for a busy reader. `data-digest` runs again and again on the *same* metrics and earns its keep by comparing this run to the last — deltas, trend direction, anomalies. If there's nothing to compare against (a single dropped file, first-ever look), use `data-brief` instead.\n\n## Steps\n\n1. **Identify the source and the cadence.** Source can be a posted CSV\u002Fspreadsheet, a SQL query result, or a connected data tool (a Composio integration like a warehouse, analytics, or BI connector when one is available). Note which it is and the period each run covers.\n2. **Find the prior baseline.** This is the whole point. Look, in order, for:\n   - the **previous digest** in prior thread context or **episodes** (Kortny's own last run is the cleanest baseline — reuse its numbers and wording);\n   - a prior period inside the same dataset (last week's rows vs this week's);\n   - **workspace facts** recording targets or expected ranges.\n   If no baseline exists, say \"first digest — no comparison yet, establishing the baseline\" and report levels only.\n3. **Compute the deltas** for each tracked metric: absolute and percent change vs the baseline, and the direction. See `references\u002Fcomparison.md` for the delta and anomaly conventions (what counts as material, how to handle small denominators, week-over-week vs same-period-last-year).\n4. **Flag anomalies** — a metric outside its usual range, a sign flip, a sudden spike or drop, a metric that stopped updating. An anomaly gets a one-line \"why this might be\" only if the data supports it; otherwise just flag it.\n5. **Lead with what changed**, not with a full table. The standout movements go first; steady metrics get a single \"holding steady\" line.\n6. **Name the action** if the deltas support one (\"retention dropped two weeks running — worth a look before it compounds\").\n\n## Output shape (Slack mrkdwn)\n\n```\n*Data digest — week of Jun 9* (vs week of Jun 2)\n\n*Moved*\n• Activation 41% → 58% (+17pts) — best week this quarter.\n• Weekly active 12.1k → 11.4k (−6%) — second week down.\n\n*Anomalies*\n• Signup→trial conversion jumped to 22% (usual ~12%) — check the source; possible double-count.\n\n*Holding steady*\n• Churn ~3%, NPS 41 — no material change.\n\n*Worth acting on:* the two-week active-user slide. Want the cohort cut?\n```\n\nOffer the table or a chart as a follow-up artifact (`chart-maker` \u002F `spreadsheet-builder`) rather than dumping it inline.\n\n## Rules\n\n- Every number is from the data. Never round a guess into a figure, and never fabricate a baseline to manufacture a delta.\n- Percent changes need a real denominator — on tiny numbers report the absolute change and say the base is small, don't post a misleading \"+300%\".\n- If the source didn't refresh (same rows as last run), say so plainly: \"source hasn't updated since last digest\" — a stale digest pretending to be fresh is worse than none.\n- Anomaly callouts are observations, not diagnoses. Speculate on cause only when the data points at it; otherwise flag and ask.\n- Keep prior-run wording stable so readers can track a metric across digests.\n- This is built to run on a schedule — pair with Kortny's scheduler for \"post this every Monday\"; the body is identical by hand or on schedule, and the scheduled run is what makes the comparison automatic.\n\n## Worked example\n\nScheduled Monday run over a metrics sheet, prior digest in episodes:\n\n> *Data digest — week of Jun 9* (vs week of Jun 2)\n>\n> *Moved*\n> • Trial→paid 18% → 24% (+6pts) — strongest since March.\n> • Weekly active 12.1k → 11.4k (−6%) — down two weeks now.\n>\n> *Anomalies*\n> • Support tickets 90 → 210 (+133%) — spike concentrated on Thursday; possible incident, not a trend.\n>\n> *Holding steady*\n> • MRR growth ~4%\u002Fwk, churn ~3%.\n>\n> *Worth acting on:* the active-user decline and Thursday's ticket spike — they may be the same story. Want me to pull the Thursday cohort?\n","## Goal\n\nProduce a recurring data readout whose value is *comparison*: what moved since last time, what's anomalous, what's worth acting on — across whatever source the data lives in.\n\nThis is the recurring, longitudinal cousin of `data-brief`. `data-brief` turns one posted file into a one-shot story for a busy reader. `data-digest` runs again and again on the *same* metrics and earns its keep by comparing this run to the last — deltas, trend direction, anomalies. If there's nothing to compare against (a single dropped file, first-ever look), use `data-brief` instead.\n\n## Steps\n\n1. **Identify the source and the cadence.** Source can be a posted CSV\u002Fspreadsheet, a SQL query result, or a connected data tool (a Composio integration like a warehouse, analytics, or BI connector when one is available). Note which it is and the period each run covers.\n2. **Find the prior baseline.** This is the whole point. Look, in order, for:\n   - the **previous digest** in prior thread context or **episodes** (Kortny's own last run is the cleanest baseline — reuse its numbers and wording);\n   - a prior period inside the same dataset (last week's rows vs this week's);\n   - **workspace facts** recording targets or expected ranges.\n   If no baseline exists, say \"first digest — no comparison yet, establishing the baseline\" and report levels only.\n3. **Compute the deltas** for each tracked metric: absolute and percent change vs the baseline, and the direction. See `references\u002Fcomparison.md` for the delta and anomaly conventions (what counts as material, how to handle small denominators, week-over-week vs same-period-last-year).\n4. **Flag anomalies** — a metric outside its usual range, a sign flip, a sudden spike or drop, a metric that stopped updating. An anomaly gets a one-line \"why this might be\" only if the data supports it; otherwise just flag it.\n5. **Lead with what changed**, not with a full table. The standout movements go first; steady metrics get a single \"holding steady\" line.\n6. **Name the action** if the deltas support one (\"retention dropped two weeks running — worth a look before it compounds\").\n\n## Output shape (Slack mrkdwn)\n\n```\n*Data digest — week of Jun 9* (vs week of Jun 2)\n\n*Moved*\n• Activation 41% → 58% (+17pts) — best week this quarter.\n• Weekly active 12.1k → 11.4k (−6%) — second week down.\n\n*Anomalies*\n• Signup→trial conversion jumped to 22% (usual ~12%) — check the source; possible double-count.\n\n*Holding steady*\n• Churn ~3%, NPS 41 — no material change.\n\n*Worth acting on:* the two-week active-user slide. Want the cohort cut?\n```\n\nOffer the table or a chart as a follow-up artifact (`chart-maker` \u002F `spreadsheet-builder`) rather than dumping it inline.\n\n## Rules\n\n- Every number is from the data. Never round a guess into a figure, and never fabricate a baseline to manufacture a delta.\n- Percent changes need a real denominator — on tiny numbers report the absolute change and say the base is small, don't post a misleading \"+300%\".\n- If the source didn't refresh (same rows as last run), say so plainly: \"source hasn't updated since last digest\" — a stale digest pretending to be fresh is worse than none.\n- Anomaly callouts are observations, not diagnoses. Speculate on cause only when the data points at it; otherwise flag and ask.\n- Keep prior-run wording stable so readers can track a metric across digests.\n- This is built to run on a schedule — pair with Kortny's scheduler for \"post this every Monday\"; the body is identical by hand or on schedule, and the scheduled run is what makes the comparison automatic.\n\n## Worked example\n\nScheduled Monday run over a metrics sheet, prior digest in episodes:\n\n> *Data digest — week of Jun 9* (vs week of Jun 2)\n>\n> *Moved*\n> • Trial→paid 18% → 24% (+6pts) — strongest since March.\n> • Weekly active 12.1k → 11.4k (−6%) — down two weeks now.\n>\n> *Anomalies*\n> • Support tickets 90 → 210 (+133%) — spike concentrated on Thursday; possible incident, not a trend.\n>\n> *Holding steady*\n> • MRR growth ~4%\u002Fwk, churn ~3%.\n>\n> *Worth acting on:* the active-user decline and Thursday's ticket spike — they may be the same story. Want me to pull the Thursday cohort?\n",{"name":5,"metadata":24,"description":7},{"tags":25,"version":26,"display_name":27},"data, digest, recurring, metrics, trends, anomalies, deltas","1.0.0","Data Digest","main","e54c3c09fd8ab6548a56c4a0e91a5af4c5e58774","kortny\u002Fskills\u002Fcurated\u002Fdata-digest",3,"https:\u002F\u002Fgithub.com\u002Fmatangilabs\u002Fkortny","https:\u002F\u002Fkortny.dev","https:\u002F\u002Fgithub.com\u002Fmatangilabs","www.matangilabs.com","Building AI coworkers that live where your team works. Makers of Kortny.","2026-10-02 13:41:52",{"source":8,"skillName":5,"commands":39,"prompt":49,"zipAvailable":50},[40,43,46],{"tool":41,"command":42},"npx","npx skills add matangilabs\u002Fkortny --skill data-digest",{"tool":44,"command":45},"bunx","bunx skills add matangilabs\u002Fkortny --skill data-digest",{"tool":47,"command":48},"pnpm","pnpm dlx skills add matangilabs\u002Fkortny --skill data-digest","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fmatangilabs\u002Fkortny\u002Fblob\u002Fmain\u002Fkortny\u002Fskills\u002Fcurated\u002Fdata-digest\u002FSKILL.md",true,[52],{"path":53,"resourceType":54,"resourceTypeDesc":55,"sizeBytes":56,"sourceHtmlUrl":57},"references\u002Fcomparison.md","REFERENCE","参考文档",2506,"https:\u002F\u002Fgithub.com\u002Fmatangilabs\u002Fkortny\u002Fblob\u002Fmain\u002Fkortny\u002Fskills\u002Fcurated\u002Fdata-digest\u002Freferences\u002Fcomparison.md",[59,66,75,82,89,96],{"skillNo":60,"name":61,"nameZh":6,"description":62,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":63,"categories":64,"keywords":65,"publishedAt":19},"01M76RCRTNAEPXFXVHARYVFY38","demo-skill","Use when the user asks to run the demo analysis workflow on a dataset.","https:\u002F\u002Fgithub.com\u002Fmatangilabs\u002Fkortny\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Fdemo-skill\u002FSKILL.md",[],[],{"skillNo":67,"name":68,"nameZh":6,"description":69,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":70,"licenseFileUrl":71,"sourceHtmlUrl":72,"categories":73,"keywords":74,"publishedAt":19},"01M76RCPW81Z7FS3FCBY67FEMQ","youtube-transcript","Use when asked to get a transcript, extract key points, or summarize a YouTube video — given a YouTube URL.","MIT","https:\u002F\u002Fgithub.com\u002Fmatangilabs\u002Fkortny\u002Fblob\u002Fmain\u002Fkortny\u002Fskills\u002Fcurated\u002Fyoutube-transcript\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fmatangilabs\u002Fkortny\u002Fblob\u002Fmain\u002Fkortny\u002Fskills\u002Fcurated\u002Fyoutube-transcript\u002FSKILL.md",[],[],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":19},"01M7547HV0ZBTKNWMMSK1CBDYQ","weekly-status-report","Use when asked to compile a weekly status update, team recap, or progress report from recent channel activity, tasks, or notes.","https:\u002F\u002Fgithub.com\u002Fmatangilabs\u002Fkortny\u002Fblob\u002Fmain\u002Fkortny\u002Fskills\u002Fcurated\u002Fweekly-status-report\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":19},"01M7547EKHDHYTPR1ZAQ2E6MN2","weekly-channel-digest","Use when asked to digest a channel's week, recap what happened in a channel, catch someone up on a channel, or post a recurring \"this week in","https:\u002F\u002Fgithub.com\u002Fmatangilabs\u002Fkortny\u002Fblob\u002Fmain\u002Fkortny\u002Fskills\u002Fcurated\u002Fweekly-channel-digest\u002FSKILL.md",[],[],{"skillNo":90,"name":91,"nameZh":6,"description":92,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":93,"categories":94,"keywords":95,"publishedAt":19},"01M7547AYJZX4N1BN7FEQC0ARQ","thread-recap","Use when asked to recap this thread, summarize the discussion above, TL;DR a long thread, or pull the decisions and action items out of the conversation right here — work from the live Slack thread in front of you and post a tight recap back into it.","https:\u002F\u002Fgithub.com\u002Fmatangilabs\u002Fkortny\u002Fblob\u002Fmain\u002Fkortny\u002Fskills\u002Fcurated\u002Fthread-recap\u002FSKILL.md",[],[],{"skillNo":97,"name":98,"nameZh":6,"description":99,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":100,"sourceHtmlUrl":101,"categories":102,"keywords":103,"publishedAt":19},"01M75477448MG98MZCQQ3JS403","theme-factory","Use when asked to create, adapt, or apply a visual theme — color palette, typography system, component tokens — for a product, document, or presentation.","https:\u002F\u002Fgithub.com\u002Fmatangilabs\u002Fkortny\u002Fblob\u002Fmain\u002Fkortny\u002Fskills\u002Fcurated\u002Ftheme-factory\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fmatangilabs\u002Fkortny\u002Fblob\u002Fmain\u002Fkortny\u002Fskills\u002Fcurated\u002Ftheme-factory\u002FSKILL.md",[],[]]