本文へ移動
cccskills
無料GitHub で公開

find-in-granola

Answer "what did we say, decide, or promise" from Granola meeting notes — narrow to the meetings that can hold the answer, read the notes before the transcript, and cite the meeting behind every claim. Use when the user asks what happened in a meeting, what a customer or teammate said, what was decided or promised, or refers to a call, sync, standup, or demo.

インストール方法を見る

含まれるファイル(1)

  • SKILL.md3.2 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Find what was actually said

Meeting notes are what people remembered, plus a summary a model wrote afterwards. The job is to locate the right meeting, read what it records, and separate what was decided from what was merely discussed.

1. Narrow to the meetings that can contain the answer

Start from list_meetings scoped by time range and involvement, not from a broad query. "The customer call last week" is a date range and an attendee, not a keyword. If the user names a folder or team space, list_meeting_folders first and list inside it. When results look empty, get_account_info tells you which email and workspace you are reading from — a personal account connected where the work one was meant is the usual cause, so say which one you searched.

2. Notes first, transcript only when the wording matters

get_meetings returns the summarized notes and any private notes, which is enough for "what was decided". Reach for get_meeting_transcript only when the exact phrasing matters — a commitment, a number, a quote the user wants verbatim — and quote the line rather than paraphrasing it. Transcripts are long; read them for the passage you need, not front to back.

3. Ask the question tool questions, not lookups

query_granola_meetings answers a natural-language question across many meetings and cites its sources. Use it when the question spans meetings ("what have customers said about pricing this quarter"). Do not use it to fetch a meeting you already identified — read that one directly, because a synthesized answer blurs which meeting said what.

4. Separate decided from discussed

Notes mix proposals, objections, and outcomes. Report a decision only when the notes record one — an action item, an "agreed", an owner and a date. If the notes show discussion without a conclusion, say the question was raised and not resolved; that is a real answer.

5. Cite the meeting

Every claim names the meeting title and date it came from, with a link when the tool returns one. Two meetings that disagree are a finding to report, not a conflict to resolve silently — the later one is not automatically right.

Anti-patterns

  • Searching by keyword when the user gave you a date or a person. Meeting titles rarely contain the topic; attendees and dates find the meeting.
  • Quoting the summary as if someone said it. Summaries are generated; only the transcript holds words people actually spoke.
  • Reading one instance of a recurring meeting. A weekly sync decides things across weeks; check the neighbours before saying something was never raised.
  • Reporting "no meetings found" without the scope. Name the workspace, folder, and time range you searched, so an empty result can be checked.

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

10x

無料

Personalized audit that teaches the advanced Superset features the user isn't using yet (automations, parallel agents, tasks, multi-host, terminal remote control, custom commands, MCP, pages) and sets them up live. Use when the user wants to get more out of Superset, asks "what else can Superset do", "how do I 10x my workflow", "what am I missing", or wants to learn a specific Superset feature.

日本語の概要は準備中です。原文の説明を表示しています。

superset-sh/superset1.5万2026年10月13日 更新

automate

無料

Turn a recurring chore into a Superset automation. Drafts the agent prompt, confirms the schedule or event trigger and the target, creates it with the CLI, and reviews the first run together. Use when the user wants a scheduled or recurring agent, a daily or weekly job, or an agent that reacts to a Slack message, GitHub pull request, Linear issue, Sentry issue or incoming email, or says things like "every morning do X", "automate this", "set up a cron agent", "run this on a schedule", "when someone opens a PR, do X".

日本語の概要は準備中です。原文の説明を表示しています。

superset-sh/superset1.5万2026年10月13日 更新

browser

無料

Open, navigate, screenshot, read, click, and type in web pages from an agent. Use when the user asks to open a URL, preview or verify a running web app, check a page's console, fill a form, click through a web flow, or automate anything in a browser, including "open localhost:3000", "screenshot the page", "what does the console say", "click the submit button", "I'm logged in on Chrome". Drives the in-app browser pane of a Superset workspace by default and Browser Use for browsers the panes cannot reach.

日本語の概要は準備中です。原文の説明を表示しています。

superset-sh/superset1.5万2026年10月13日 更新

Write Stripe code against what the API does today — plan the product shape, confirm every parameter from Stripe's own docs and schemas, and keep the agent in a sandbox. Use when adding checkout, billing, subscriptions, invoicing, or webhooks to a codebase, or when debugging Stripe integration code that compiles but behaves wrong.

日本語の概要は準備中です。原文の説明を表示しています。

superset-sh/superset1.5万2026年10月13日 更新

Verify UI behavior end-to-end by driving the running desktop app over the Chrome DevTools Protocol. Use when asked to verify, reproduce, or confirm a UI change, bug, or regression in the real app rather than in tests.

日本語の概要は準備中です。原文の説明を表示しています。

superset-sh/superset1.5万2026年10月13日 更新

ci-triage

無料

Diagnose a failing GitHub Actions run — find the first real error in the logs, tell a flake apart from a genuine failure, and identify the commit that broke it. Use when checks are red on a PR, a workflow failed, the build is broken, CI is flaky, or the user asks why a run failed or whether a failure is real.

日本語の概要は準備中です。原文の説明を表示しています。

superset-sh/superset1.5万2026年10月13日 更新

superset-sh のスキルをすべて見る

このスキルの問題を報告する