Granola Reference ArchitectureSAFE
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-reference-architecture description: 'Enterprise reference architecture for meeting management with Granola. Use when designing org-wide meeting workflows, planning integration topology, or architecting meeting-to-action pipelines across departments. Trigger: "granola architecture", "granola enterprise design", "granola system design", "meeting system architecture". ' allowed-tools: Read, Write, Edit version: 1.13.0 license: MIT author: Jeremy Longshore <[email protected]> tags: - saas - granola - architecture - enterprise compatibility: Designed for Claude Code --- # Granola Reference Architecture ## Overview Enterprise reference architecture for deploying Granola as the meeting intelligence platform across an organization. Covers the core capture pipeline, Zapier middleware routing, multi-workspace topology, security layers, and integration patterns for Slack, Notion, CRM, and task management. ## Prerequisites - Granola Enterprise plan ($35+/user/month) - Zapier Professional or higher (for multi-step Zaps and Paths) - Destination systems provisioned (Slack, Notion, CRM, Linear/Jira) - IT architecture review completed ## Instructions ### Step 1 — Core Pipeline Architecture ``` ┌─────────────────────────────────────────────────────────┐ │ MEETING PLATFORMS │ │ Zoom | Google Meet | Microsoft Teams | Slack Huddle│ └──────────────────────┬──────────────────────────────────┘ │ System audio capture (no bot) ▼ ┌─────────────────────────────────────────────────────────┐ │ GRANOLA PLATFORM │ │ │ │ ┌──────────┐ ┌──────────────┐ ┌─────────────────┐ │ │ │ Audio │→ │ Transcription │→ │ AI Enhancement │ │ │ │ Capture │ │ (GPT-4o/ │ │ (Notes + Trans- │ │ │ │ (local) │ │ Claude) │ │ cript merge) │ │ │ └──────────┘ └──────────────┘ └────────┬────────┘ │ │ │ │ │ ┌────────────────┐ ┌──────────────────┐ │ │ │ │ People & │ │ Folders │ │ │ │ │ Companies │ │ (routing rules) │←─┘ │ │ │ (built-in CRM) │ └────────┬─────────┘ │ │ └────────────────┘ │ │ └───────────────────────────────┼─────────────────────────┘ │ ┌───────────────────┼───────────────────┐ │ ZAPIER MIDDLEWARE │ │ │ │ ┌─────────┐ ┌─────────┐ ┌──────┐ │ │ │ Filter │→ │ Route │→ │ Act │ │ │ │ (type/ │ │ (paths │ │ │ │ │ │ attendee)│ │ /rules) │ │ │ │ │ └─────────┘ └─────────┘ └──────┘ │ └──────┬──────────┬──────────┬─────────┘ │ │ │ ┌──────────┴──┐ ┌────┴────┐ ┌───┴────────┐ │ COMMUNICATE │ │ ARCHIVE │ │ ACT │ │ Slack │ │ Notion │ │ Linear/Jira│ │ Email │ │ Drive │ │ Asana │ │ Teams │ │ GCS/S3 │ │ HubSpot │ └─────────────┘ └─────────┘ └────────────┘ ``` ### Step 2 — Folder-Based Routing Topology Design folder structure to drive automated routing: | Folder | Trigger Type | Destinations | Permissions | |--------|-------------|-------------|-------------| | `Sales / Discovery` | Auto | Slack #sales + HubSpot Deal + Follow-up Email | Sales team | | `Sales / Demo` | Auto | Slack #sales + HubSpot Contact | Sales team | | `Engineering / Sprint` | Auto | Slack #eng + Linear tasks + Notion wiki | Engineering | | `Engineering / Architecture` | Auto | Slack #eng-arch + Notion ADR database | Senior engineers | | `Product / Customer Feedback` | Auto | Slack #product + Notion feedback DB | Product team | | `Leadership / All-Hands` | Auto | Slack #general + Google Drive archive | All | | `Leadership / Board` | Manual only | Private Notion, no Slack | Executives only | | `HR / Interviews` | Manual share | Greenhouse scorecard | Hiring managers | | `HR / 1-on-1s` | None | Private, no automation | Individual | ### Step 3 — Multi-Workspace Topology ``` Organization (Enterprise) ├── Engineering Workspace │ ├── Folders: Sprint, Architecture, Standup, Retro │ ├── Integrations: Linear, Notion, Slack #engineering │ └── Retention: 2 years notes, 90 days transcripts │ ├── Sales Workspace │ ├── Folders: Discovery, Demo, Pipeline Review │ ├── Integrations: HubSpot, Slack #sales, Gmail │ └── Retention: 1 year notes, 90 days transcripts │ ├── Product Workspace │ ├── Folders: Customer Feedback, Design Review, PRD │ ├── Integrations: Notion, Linear, Slack #product │ └── Retention: 2 years notes, 90 days transcripts │ ├── HR Workspace (Confidential) │ ├── Folders: Interviews, 1-on-1s, Performance │ ├── Integrations: Greenhouse (via Zapier) │ ├── Retention: 30 days notes, 7 days transcripts │ └── Restrictions: No external sharing, MFA required │ └── Executive Workspace (Confidential) ├── Folders: Board, Strategy, M&A ├── Integrations: Private Notion only ├── Retention: Custom (legal hold capable) └── Restrictions: IP allowlist, 4-hour session timeout ``` ### Step 4 — Security Architecture ``` Authentication Layer: SSO (Okta/Azure AD) → SCIM provisioning → JIT workspace assignment Data Protection: In Transit: TLS 1.3 (all API calls) At Rest: AES-256 (server-side storage) Local: FileVault/BitLocker (cache-v3.json encryption) Access Control: Org Owner → Workspace Admin → Team Lead → Member → Viewer → Guest ↓ SSO groups mapped to Granola roles per workspace Data Classification: Confidential: HR, Legal, Board (restricted sharing, short retention) Internal: Engineering, Product (team sharing, standard retention) External-Facing: Sales, Cus
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-reference-architecture.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 Reference Architecture 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 Reference Architecture 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 Reference Architecture access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
Which assistants does Granola Reference Architecture 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.