# AGENTS.md — Build guide for Granola

## Project scope
Record consented microphone/system audio with an always-visible status, transcribe locally and place the transcript beside the user's rough notes. Show the exact text before any optional remote summarization request.

Catalogue verdict: yes. A local recorder plus Whisper transcription plus an LLM note template covers the core personal loop in one sitting. The paid product is mainly polish, meeting context, sync, and team workflow.
Use the implementation prompt below to define the deliverable. Complete each phase's acceptance checks before extending the scope.

## Working agreement
- Inspect the repository and its existing instructions before choosing paths, dependencies or commands. Keep one coherent stack and explain changes to the proposed architecture.
- Plan a vertical slice that accepts a real input and produces the useful output described below. Persist only the state the prompt calls for; respect memory-only and upstream-managed workflows. Use fixtures only when they are clearly labelled.
- After scaffolding, document the actual install, development, check and build commands in README and keep them synchronized with the package or project manifest. Do not report commands as successful unless they ran.
- Work in small steps. At handoff, list implemented flows, checks actually performed, remaining blockers, and any credentials or provider setup the owner must supply.
- Do not publish, spend money, contact customers, delete source data or run irreversible migrations without the project owner's authorization.

## Prerequisites
- macOS, Xcode and a selected deployment target. Ask for each OS permission only when its feature is enabled; code signing and distribution are separate release tasks. Screen/microphone permission when recording and a downloaded compatible local transcription model.
- Implementation components: Swift, SwiftUI and focused AppKit integrations for a macOS utility. SQLite or Codable files for local settings and history; macOS Keychain for credentials. ScreenCaptureKit/AVFoundation recording and a documented local speech-model executable; optional remote summary adapter only after transcript approval.
- Scope boundary: sync across devices; the mobile app

## Stack and architecture
- Swift, SwiftUI and focused AppKit integrations for a macOS utility.
- SQLite or Codable files for local settings and history; macOS Keychain for credentials.
- ScreenCaptureKit/AVFoundation recording and a documented local speech-model executable; optional remote summary adapter only after transcript approval.
- Domain model: recording sessions, local audio tracks, rough notes, timestamped transcript segments and accepted summaries

## Security and data integrity
- Start recording only after explicit microphone/system-audio selection and consent. Show a persistent recording indicator and stop control, document retention and keep raw audio/notes out of logs. Preview selected transcript text before any remote summary request; no keystroke or clipboard monitoring is needed.
- Correctness boundary: The authored notes are never overwritten; recording permission is explicit and remote processing requires a separate action.
- Keep UI state on the main actor, cancel obsolete background tasks, and avoid blocking system callbacks. Record only the selected feature scope, with clear retention controls.
- Export settings and explicitly selected history without secrets. Preserve the last working configuration and provide a reset that does not delete user source files.

## Agent implementation rules
- Project rule — data model: recording sessions, local audio tracks, rough notes, timestamped transcript segments and accepted summaries
- Project rule — preserve this invariant: The authored notes are never overwritten; recording permission is explicit and remote processing requires a separate action.
- Project rule — acceptance evidence: Deny microphone access and create no empty successful recording; reject a summary and keep raw notes and audio, with every accepted quotation linked to a transcript segment.

## Optional agent skills and references
- Optional external skill: [swiftui-expert-skill](https://github.com/AvdLee/SwiftUI-Agent-Skill/blob/main/skills/swiftui-expert-skill/SKILL.md) — Build and review native SwiftUI views, state management, navigation, accessibility and rendering performance. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.
- Optional external skill: [swift-concurrency](https://github.com/AvdLee/Swift-Concurrency-Agent-Skill/blob/main/skills/swift-concurrency/SKILL.md) — Design Swift async tasks, actors, isolation, cancellation and safe data sharing, including Swift 6 migration. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.
- Optional external skill: [sharp-edges](https://github.com/trailofbits/skills/blob/main/plugins/sharp-edges/skills/sharp-edges/SKILL.md) — Review security-sensitive APIs and configuration for dangerous defaults and easy-to-misuse interfaces. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.

Read the linked SKILL.md and its dependencies before adding a skill. Select only the skills matching this project's runtime and task; their documentation does not supply API access, credentials or approval to perform external actions. Pin the reviewed revision where the tool supports it. Follow the chosen agent's documented project-level installation mechanism.

## Distribution ideas
These are optional planning notes. Obtain the owner's approval before publishing or contacting anyone.
- Demonstrate the actual Granola-inspired workflow with owned or clearly labeled sample data: Record consented microphone/system audio with an always-visible status, transcribe locally and place the transcript beside the user's rough notes. Show the exact text before any optional remote summarization request.
- Publish a reproducible walkthrough with this observable result: Deny microphone access and create no empty successful recording; reject a summary and keep raw notes and audio, with every accepted quotation linked to a transcript segment.
- Explain who can operate this scoped tool, its setup and ongoing costs, and these remaining product gaps: sync across devices; the mobile app Avoid guaranteed savings, performance scores or implied endorsement.

## Engineering roadmap
1. Phase 1 — Scope and fixtures. Implement this bounded workflow: Record consented microphone/system audio with an always-visible status, transcribe locally and place the transcript beside the user's rough notes. Show the exact text before any optional remote summarization request. Record prerequisites, select representative user-owned fixtures and document the unsupported features: sync across devices; the mobile app
2. Phase 2 — Durable model. Model recording sessions, local audio tracks, rough notes, timestamped transcript segments and accepted summaries Add migrations or a versioned document format, explicit validation, stable IDs and a visible import-error report. Preserve this rule: The authored notes are never overwritten; recording permission is explicit and remote processing requires a separate action.
3. Phase 3 — Complete the first useful path. Implement the workflow's input, review and output interface, with clear controls and explicit empty/error states. Keep UI state on the main actor, cancel obsolete background tasks, and avoid blocking system callbacks. Record only the selected feature scope, with clear retention controls.
4. Phase 4 — Permissions and integration failure. Start recording only after explicit microphone/system-audio selection and consent. Show a persistent recording indicator and stop control, document retention and keep raw audio/notes out of logs. Preview selected transcript text before any remote summary request; no keystroke or clipboard monitoring is needed. Request integration credentials and permissions only for the enabled feature; show a disconnected state instead of mock results.
5. Phase 5 — Portable handoff. Export settings and explicitly selected history without secrets. Preserve the last working configuration and provide a reset that does not delete user source files. Include setup, operating limits, fixture walkthrough and shutdown/restart instructions in the README.
6. Phase 6 — Acceptance scenarios. Deny microphone access and create no empty successful recording; reject a summary and keep raw notes and audio, with every accepted quotation linked to a transcript segment. Repeat the workflow after restart and with a denied permission or unavailable dependency; show recoverable failure rather than a success placeholder.

## Paid-product capabilities outside this build
- sync across devices
- the mobile app
- automatic calendar joins
- someone else's servers & support

## Implementation prompt
WORKING SLICE
Record consented microphone/system audio with an always-visible status, transcribe locally and place the transcript beside the user's rough notes. Show the exact text before any optional remote summarization request.

Build this scoped Granola-inspired workflow with a documented data model and visible failure states.

Architecture
- Swift, SwiftUI and focused AppKit integrations for a macOS utility.
- SQLite or Codable files for local settings and history; macOS Keychain for credentials.
- ScreenCaptureKit/AVFoundation recording and a documented local speech-model executable; optional remote summary adapter only after transcript approval.

Prerequisites and limits
macOS, Xcode and a selected deployment target. Ask for each OS permission only when its feature is enabled; code signing and distribution are separate release tasks. Screen/microphone permission when recording and a downloaded compatible local transcription model.
Outside this release: sync across devices; the mobile app

Data model and correctness
recording sessions, local audio tracks, rough notes, timestamped transcript segments and accepted summaries
Invariant: The authored notes are never overwritten; recording permission is explicit and remote processing requires a separate action.
Keep UI state on the main actor, cancel obsolete background tasks, and avoid blocking system callbacks. Record only the selected feature scope, with clear retention controls.

Security and privacy
Start recording only after explicit microphone/system-audio selection and consent. Show a persistent recording indicator and stop control, document retention and keep raw audio/notes out of logs. Preview selected transcript text before any remote summary request; no keystroke or clipboard monitoring is needed.

Recovery and export
Export settings and explicitly selected history without secrets. Preserve the last working configuration and provide a reset that does not delete user source files.

Implementation order
1. Phase 1 — Scope and fixtures. Implement this bounded workflow: Record consented microphone/system audio with an always-visible status, transcribe locally and place the transcript beside the user's rough notes. Show the exact text before any optional remote summarization request. Record prerequisites, select representative user-owned fixtures and document the unsupported features: sync across devices; the mobile app
2. Phase 2 — Durable model. Model recording sessions, local audio tracks, rough notes, timestamped transcript segments and accepted summaries Add migrations or a versioned document format, explicit validation, stable IDs and a visible import-error report. Preserve this rule: The authored notes are never overwritten; recording permission is explicit and remote processing requires a separate action.
3. Phase 3 — Complete the first useful path. Implement the workflow's input, review and output interface, with clear controls and explicit empty/error states. Keep UI state on the main actor, cancel obsolete background tasks, and avoid blocking system callbacks. Record only the selected feature scope, with clear retention controls.
4. Phase 4 — Permissions and integration failure. Start recording only after explicit microphone/system-audio selection and consent. Show a persistent recording indicator and stop control, document retention and keep raw audio/notes out of logs. Preview selected transcript text before any remote summary request; no keystroke or clipboard monitoring is needed. Request integration credentials and permissions only for the enabled feature; show a disconnected state instead of mock results.
5. Phase 5 — Portable handoff. Export settings and explicitly selected history without secrets. Preserve the last working configuration and provide a reset that does not delete user source files. Include setup, operating limits, fixture walkthrough and shutdown/restart instructions in the README.
6. Phase 6 — Acceptance scenarios. Deny microphone access and create no empty successful recording; reject a summary and keep raw notes and audio, with every accepted quotation linked to a transcript segment. Repeat the workflow after restart and with a denied permission or unavailable dependency; show recoverable failure rather than a success placeholder.

Acceptance
Deny microphone access and create no empty successful recording; reject a summary and keep raw notes and audio, with every accepted quotation linked to a transcript segment.
Use real source data or clearly labeled fixtures. Explain unsupported input and provider failures; do not fabricate analytics, delivery receipts, accuracy claims or security guarantees.

Optional agent guidance
Optional external skill: [swiftui-expert-skill](https://github.com/AvdLee/SwiftUI-Agent-Skill/blob/main/skills/swiftui-expert-skill/SKILL.md) — Build and review native SwiftUI views, state management, navigation, accessibility and rendering performance. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.
Optional external skill: [swift-concurrency](https://github.com/AvdLee/Swift-Concurrency-Agent-Skill/blob/main/skills/swift-concurrency/SKILL.md) — Design Swift async tasks, actors, isolation, cancellation and safe data sharing, including Swift 6 migration. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.
Optional external skill: [sharp-edges](https://github.com/trailofbits/skills/blob/main/plugins/sharp-edges/skills/sharp-edges/SKILL.md) — Review security-sensitive APIs and configuration for dangerous defaults and easy-to-misuse interfaces. Review its instructions and compatibility before use; it does not grant deployment, data-access or publication permission.
Project rule — data model: recording sessions, local audio tracks, rough notes, timestamped transcript segments and accepted summaries
Project rule — preserve this invariant: The authored notes are never overwritten; recording permission is explicit and remote processing requires a separate action.
Project rule — acceptance evidence: Deny microphone access and create no empty successful recording; reject a summary and keep raw notes and audio, with every accepted quotation linked to a transcript segment.

## Completion evidence
Demonstrate the prompt's acceptance scenarios against the scoped workflow. Include setup from a clean checkout and failure recovery. Check persistence across restart and export/restore only for the state the prompt says to store; for memory-only tools, confirm that temporary content is discarded as specified. Record actual results and remaining limitations. A detailed plan alone does not establish a working replacement.