Granola Makes Repeatable Prompts With Recipes

Granola, the AI note-taking app for meetings, is launching Recipes, a feature that allows repeatable prompts to be turned into reusable workflows, so one-off chat instructions will now work with Granola. The update aims to address a familiar pain point: repeating the same analyses and follow-ups after every call, as well as needing to retype prompts on each occasion.

What Granola Recipes are and how users apply them

Recipes are saved prompt shortcuts that can be used in Granola chat by writing the prompt name prefixed with “/”. Users can create their own prompts, assign them context in a single meeting or across meetings, and save them to share with teammates.

Granola also ships with a curated library of prebuilt conveniences organized around the meeting lifecycle itself—before, during and after. Think pre-call briefings, real-time agenda reinforcement and structured debriefs that consistently bring in action items, risks or customer objections—without reinventing the prompt with every plan.

How repeatable prompts work inside Granola’s Recipes

In reality, a user might type “/qbr-outline” to generate a quarterly business review skeleton from a pipeline review, or “/risks” to export blockers and mitigation owners at the end of a standup.

The point is that since the recipe knows the context of the meeting, it can refer to speakers, decisions and previous notes rather than parsing a raw transcript from scratch.

Recipes can be scoped. Meeting-specific one-off recipes are bound to a session only; cross-meeting recipes survive in future calls and may be invoked at any time for the same kind of meeting. Teams could use shared recipes for publishing so that sales, success, product or recruiting each have a consistent prompt playbook.

For users who want guardrails, Granola provides a recipe for building effective recipes—detailing prompt structure, context cues and output formatting. That makes a difference—because prompt drift is a real thing: without an authoritative instruction, even mighty models will spit different answers.

Why This Is Important For Productivity in Meetings

Knowledge workers spend a huge share of their time in meetings—and even more time, often after hours, cleaning up the mess. Managers spend an average of about 23 hours a week in meetings, Harvard Business Review has found, while Microsoft’s Work Trend Index has found that meeting time has exploded since 2020. It’s not just that friction is costly; it’s the constant summarizing, tracking and follow-through.

Repeatable prompts move that effort from ad hoc to programmatic. Rather than “summarize the call” in 10 different ways across a team, a standard template produces the same structure every instance—headlines, decision items with owners and due dates, etc.—so that the outputs fit neatly into CRM notes, Jira items or internal docs.

There are two practical advantages to this: it means meeting outcomes can be processed faster, and the results from different sessions will be more comparable. With the same fields on every discovery call, it’s easy for leaders to see commonalities—objections that come up again and again; gaps compared to competitors or other offerings; hand-offs that are ending in dropped balls—without trudging through freeform notes.

Competition and differentiation among AI meeting apps

AI tools more broadly are having a moment with repeatable prompts. Chatbot platforms ChatGPT and Poe have made bespoke GPTs and bots popular; newer browsers like Dia and Opera Neon feature card- or skill-like automations to execute a bunch of tasks you need done on the regular. And in the meeting space itself, Fireflies, Fathom and Circleback provide templates and prompt-driven insights typically done post-meeting.

The pitch of Granola is timing and context. By integrating recipes into the workflow of the live meeting—and allowing users to call them up via lightweight slash commands—the app is aiming to chip away at that divide between conversation and structured output. The before/during/after framing also gets to automation earlier, such as by asking for an agenda or call plan before anyone joins.

Data context and integration strategy for Recipes

As Chris Pedregal, the co-founder of the company, tells it today, the recipe in question is actually just “the app itself”—which includes meeting notes, transcripts and chat. The company is expected to link more services together so recipes can read from and write back to external systems, extending what a single command may accomplish.

Then it isn’t even pointing in the right direction toward how teams actually work: insights have to land where work is tracked. A sales recipe that tags MEDDICC fields is most useful when it modifies the opportunity record; a product recipe that rolls up bugs should create Jira tickets with owners and due dates. Look for early integrations to focus on CRMs, project trackers and knowledge bases.

Companies will also be looking for strong controls. As more prompts become shareable assets, it’s inevitable that admins will demand permissioning, audit trails, retention policies and redaction options—gimmes in mature SaaS governance, referenced by something like Gartner, but still a rarity when establishing security. Predictable outputs are only valuable if they’re coupled with consistent data handling.

Real-world use cases for Granola’s repeatable prompts

“/candidate-brief” before the interview will pull in a resume, role requirements and tailored questions; “/fit-summary” after the call captures competencies, concerns, go/no-go recommendation. A customer success manager might fire “/renewal-risk” mid-review as a way to flag churn signals in sentiment and usage trends called out on the call.

For product teams, “/decision-log” can pull trade-offs and owners out from design reviews, and “/retro-themes” organizes post-mortem notes into themes with example incidents. The common denominator: transient, high-value prompts that are well served by a dense meeting context.

The bottom line on Recipes and meeting productivity

Recipes bring Granola out from the realm of generic summarization and into a configurable system of record for how teams run meetings. The feature, by standardizing prompts and making them callable in the moment, is intended to reduce busywork and lift the quality of what gets made. With planned integrations, if those make it on schedule, repeatable prompts could be a kind of connective tissue between live conversation and the tools where work gets done.