Granola Hello WorldSAFE
Model-agnostic agent-skills platform with a harness-free canonical layer, verified adapters, and the ccpi package manager. Explore at tonsofskills.com.
Overview
Model-agnostic agent-skills platform with a harness-free canonical layer, verified adapters, and the ccpi package manager. Explore at tonsofskills.com.
4f83675ca38aOBSERVED · 2026-10-08Host compatibility
What the documentation claims. We have not run a compatibility test.
| Host | Status | Notes |
|---|---|---|
| claude-code | mentioned |
What it tells the agent
The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.
--- name: granola-hello-world description: 'Capture your first meeting with Granola and review AI-enhanced notes. Use when testing Granola setup, learning the notepad + transcript flow, or understanding how Enhance Notes and Granola Chat work. Trigger: "granola hello world", "first granola meeting", "try granola", "granola test". ' allowed-tools: Read, Write, Edit, Bash(pgrep:*), Bash(open:*) version: 1.13.0 license: MIT author: Jeremy Longshore <[email protected]> tags: - saas - granola - getting-started compatibility: Designed for Claude Code --- # Granola Hello World ## Overview Capture your first meeting with Granola and learn the three-phase workflow: live notepad, Enhance Notes, and Granola Chat. Granola captures audio from your device (no bot joins), transcribes it, and merges your typed notes with the full transcript to produce structured output. ## Prerequisites - Completed `granola-install-auth` setup - Calendar connected with an upcoming meeting (or create a test one) - Microphone + Screen & System Audio permissions granted ## Instructions ### Step 1 — Start a Meeting 1. Join any video call — Zoom, Google Meet, Teams, Slack Huddle, or WebEx 2. Granola auto-detects the meeting from your synced calendar 3. A floating notepad appears showing the meeting title and attendees 4. If auto-detection fails, click the Granola icon in your menu bar and select **Start Recording** ### Step 2 — Take Live Notes in the Notepad During the meeting, type rough notes directly in the Granola notepad: ```markdown - discussed Q1 priorities - sarah owns onboarding revamp - need design review by thursday - budget question - follow up with finance ``` These notes give Granola context. The AI uses your notes **plus** the full transcript to produce better output — you do not need to capture everything. ### Step 3 — End Meeting and Enhance 1. End your video call 2. Granola processes the transcript (typically 1-2 minutes) 3. Click **Enhance Notes** — Granola merges your notes with the transcript 4. Select a template (or use the default) to structure the output ### Step 4 — Review Enhanced Output Granola produces structured notes based on your template: ```markdown # Team Standup — March 22, 2026 ## Summary Discussed Q1 priorities. Agreed to focus on customer onboarding improvements. Sarah will lead the initiative with a design review by Thursday. ## Key Decisions - Onboarding revamp is the top priority for Q1 - Budget approval needed from Finance before expanding scope ## Action Items - [ ] @sarah: Schedule design review meeting by Thursday - [ ] @mike: Create onboarding improvement tickets in Linear - [ ] @alex: Follow up with Finance on budget allocation ## Participants Sarah Chen, Mike Johnson, Alex Kim ``` ### Step 5 — Use Granola Chat After enhancement, use the chat panel to ask questions about the meeting: - "List my action items" - "Draft a follow-up email to the attendees" - "What were the open questions?" - "Summarize the budget discussion" Granola Chat has full context of the transcript and your notes. ### Step 6 — Use Recipes for Repeatable Prompts Type `/` in Granola Chat to invoke Recipes — saved prompts for common tasks: - `/standup` — extract blockers and progress updates - `/follow-up` — draft a follow-up email - `/action-items` — list all action items with owners - Create custom recipes for your team's workflows ## Output - Complete meeting notes with AI-generated summary - Key decisions and action items extracted with owners - Full searchable transcript (expandable in the note) - Granola Chat available for post-meeting queries ## Tips for Better Notes 1. **Type key points during the meeting** — even rough notes dramatically improve output 2. **State names when assigning actions** — "Action item: Sarah will schedule the review" 3. **Use explicit action language** — "We decided...", "Next step is..." 4. **Summarize decisions verbally** — the AI picks up on decision language 5. **One speaker at a time** — reduces transcription errors ## Error Handling | Error | Cause | Fix | |-------|-------|-----| | No transcript generated | Wrong audio device selected | Check System Settings > Sound > Input matches your meeting audio | | Meeting not detected | Calendar event missing video link | Add Zoom/Meet/Teams link to event, or manually start recording | | Processing stuck > 5 min | Connectivity issue | Check internet, restart Granola (right-click icon > Restart) | | Empty enhanced notes | Meeting < 2 minutes | Granola needs sufficient audio for meaningful enhancement | | Poor transcription | Background noise or crosstalk | Use headset, mute when not speaking | ## People & Companies (Automatic) After your first meeting, check the **People** and **Companies** views in the sidebar. Granola automatically: - Creates contact entries from calendar attendees - Enriches with job titles, profile photos, and company info - Links all past meeting notes to each person - Lets you review full conversation history before follow-ups ## Examples Use a fictitious meeting with no real attendees and record `fixture=hello-meeting-v1; integration=sandbox; result=accepted; destination=approved; retention=none; cleanup=complete`. ## Resources - How Transcription Works - [Customize Notes with Templates](https://docs.granola.ai/help-center/taking-notes/customise-notes-with-templates) - [Granola Chat & Recipes](https://www.granola.ai/blog/say-hello-to-recipes) - [People and Companies](https://docs.granola.ai/help-center/people-and-companies) ## Next Steps Proceed to `granola-core-workflow-a` for meeting template setup and preparation workflows.
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- none-observed
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (0)
No findings outside the package's declared scope.
Gates applied: no_behavioural_pass.
4f83675ca38afull audit observations/trust-audit/skill/jeremylongshore__granola-hello-world.json · Report an issue / request a re-scanAudit history
Every audit this skill has had.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-08 | 4f83675ca38a | SAFE | B | 89 | first audit |
Questions
What does the Granola Hello World skill do?
Model-agnostic agent-skills platform with a harness-free canonical layer, verified adapters, and the ccpi package manager. Explore at tonsofskills.com.
Is Granola Hello World safe to install?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.
What can Granola Hello World access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
Which assistants does Granola Hello World work with?
Its documentation mentions claude-code. That is what the text claims, not a compatibility test we ran.
How current is this page?
The grade is for one exact copy of the source (4f83675ca38a), read on 2026-10-08. The repository is watched, and a new audit runs when it changes — this is the first audit.