Skip to content

Latest commit

 

History

History
154 lines (132 loc) · 9.21 KB

File metadata and controls

154 lines (132 loc) · 9.21 KB

Repo Layout

This is the canonical map of the live repo surface on main.

Root Contract

The repo root should only expose:

  • live product code and assets
  • canonical build and verification entry points
  • public project policy docs
  • clearly marked historical/archive zones

If a file or folder does not fit one of those jobs, it should usually live under scripts/, Tools/, docs/, or archive/.

When a root shell command is part of the public repo surface, prefer a thin wrapper at the root and keep the implementation under scripts/.

Main Commands

Use these as the active command surface:

bash scripts/dev/agent-preflight.sh
bash build-deps.sh
bash build.sh --no-open
bash build-beta.sh '' <user>
bash run-tests.sh
bash run-integration-smoke.sh
bash run-e2e-smoke.sh
bash run-slow-pasteback-smoke.sh
bash run-live-capture-smoke.sh
bash run-daily-audio-reliability.sh
python3 scripts/ops/release-gate-report.py
bash scripts/ops/transcripted-qa-bench.sh --mode quick
bash scripts/ops/transcripted-qa-bench.sh --mode full
bash scripts/ops/transcripted-qa-bench.sh --mode ui
bash scripts/ops/transcripted-qa-bench.sh --mode sparkle-update
bash scripts/ops/transcripted-qa-bench.sh --mode packaged
bash scripts/ops/transcripted-qa-bench.sh --mode corpus
bash scripts/ops/transcripted-qa-bench.sh --mode corpus-compare
swift test

Command ownership:

  • scripts/dev/agent-preflight.sh — agent preflight and suggested verification map for the current branch
  • build-deps.sh — thin root wrapper for the dependency build entrypoint
  • build.sh — thin root wrapper for the authoritative local app build; use --no-open for agent verification
  • build-beta.sh — thin root wrapper for signed beta/distribution builds
  • run-tests.sh — thin root wrapper for curated fast tests
  • run-integration-smoke.sh — thin root wrapper for app/core smoke verification
  • run-e2e-smoke.sh — thin root wrapper for deterministic release-critical artifact smoke
  • run-slow-pasteback-smoke.sh — thin root wrapper for the deterministic fake slow Cmd+V pasteback target smoke
  • run-live-capture-smoke.sh — thin root wrapper for local hardware/TCC capture smoke
  • run-daily-audio-reliability.sh — thin root wrapper for the interactive and synthetic daily audio reliability check
  • scripts/ops/release-gate-report.py — single pre-merge/release report covering QA bench, telemetry, release surfaces, and local log warnings
  • scripts/ops/transcripted-qa-bench.sh — orchestrated QA tester pass with local report output, including --mode ui for the Accessibility-driven onboarding/menu bar/Home/Settings smoke, --mode sparkle-update for fake-state Sparkle update UI proof, and --mode packaged for no-publish package smoke
  • scripts/ops/validate-meeting-corpus.py — local-only meeting corpus validator for Downloads fixtures
  • scripts/ops/compare-meeting-corpus.py — local-only Transcripted-vs-Zoom corpus comparator for Downloads fixtures
  • swift testTranscriptedCore package seam tests

For helper and legacy scripts, see scripts/README.md.

Directory Map

  • .agents/ — machine-readable agent maps for path verification and QA gates
  • .agent-review/ — sanitized review evidence for agent PRs, not current UI truth
  • .github/ — issue templates, PR template, and repository workflows
  • Sources/ — macOS app target
  • Sources/Accessibility/ — AX helpers for overlay positioning
  • Sources/Capture/ — physical dictation trigger capture and meeting hotkey routing
  • Sources/Dictation/ — dictation persistence
  • Sources/Meeting/ — app-side meeting bridge into TranscriptedCore
  • Sources/Observability/ — analytics, crash reporting, debug logging, and Sparkle updater
  • Sources/Reliability/ — wake/sleep recovery
  • Sources/Speech/ — local STT engines, router, and audio recovery
  • Sources/Support/ — shared app utilities such as paths, permissions, hotkeys, and constants
  • Sources/TranscriptedCore/ — reusable meeting transcription library
  • Sources/UI/ — app-facing UI grouped into Overlay/, MenuBar/, Settings/, and Shared/
  • Tests/ — fast tests, package tests, and integration smoke sources
  • Tools/ — standalone sibling packages; see Tools/README.md
  • docs/ — live project docs
  • docs/strategy/ — dated strategy syntheses and deep dives for product, market, and architecture planning
  • archive/ — historical code and legacy tooling kept out of the live product surface
  • config/ — app config artifacts including entitlements and nightly security manifests
  • Casks/ — committed Homebrew cask release surface
  • Resources/ — bundled app assets
  • scripts/entrypoints/ — implementations behind the thin root command wrappers

Dated audit and autoeval docs in docs/ are point-in-time evidence. Use the current command map, local CLAUDE.md, and .agents/test-matrix.yml for live instructions unless a dated doc is explicitly the target of the task.

Docs Map

Use these docs for these jobs:

  • README.md — public product overview and quick start
  • AGENT_START.md — short agent entrypoint
  • CONTRIBUTING.md — contributor setup and contribution norms
  • AGENTS.md — Codex-specific workflow rules
  • WORKFLOW.md - local GitHub Issues to Codex agent workflow contract
  • .github/ — GitHub issue templates, PR checklist, and workflow automation
  • CLAUDE.md — Claude-specific repo orientation
  • docs/agent-onboarding.md — how to interpret the repo’s doc layers
  • docs/activation-lane.md — saved Markdown, agent payoff, and return-use routing
  • docs/agent-closeout.md — compact coordinator and agent handoff format
  • docs/agent-connect.md — saved-folder and MCP handoff guidance for agents
  • docs/docs.md - documentation tone, drift checks, and follow-up PR rules
  • docs/agent-issue-orchestration.md - how to queue GitHub issues for the local Codex runner
  • docs/install-attribution-map.md — anonymous website/download/install to first-value signal map
  • docs/ops-credentials.md — Sentry, PostHog, GitHub, and Cloudflare credential lanes
  • docs/retention-cohort-analytics.md — privacy-safe PostHog habit and retention report for first/second artifact, next-day and 7-day return, repeat use, 3-days-this-week, and health-skill output
  • docs/storage-paths.md — canonical storage and fallback path map
  • docs/audio-reliability-daily-check.md — daily manual audio reliability loop and evidence contract
  • docs/qa-test-bench.md — orchestrated QA tester bench for quick, deep, UI, Sparkle update, packaged, corpus, corpus-compare, live, artifact, and synthetic audio passes
  • docs/test-automation-strategy.md — agent-first QA coverage map, gate strategy, and automation roadmap
  • docs/qa-issue-500-meeting-audio.md — manual WebRTC / meeting-volume QA matrix for issue #500
  • docs/release-packaging.md — release packaging flow
  • docs/sparkle-updates.md — Sparkle update contract
  • docs/qa-parakeet-start-failure-smoke.md — manual Parakeet start-failure regression checklist
  • docs/capture-format.md — authoritative spec for the saved Markdown capture format
  • docs/observability.md — the five diagnostic sinks and which one owns what
  • docs/privacy-first-observability.md — the three observability lanes and their privacy contract
  • docs/analytics-taxonomy-merge.md — why the analytics registry uses union merge, and its trap
  • docs/speaker-recognition-metrics.md — how speaker-recognition accuracy is measured
  • docs/speaker-eval-exemplar-delta-2026-07.md — dated multi-exemplar / negative-exemplar accuracy eval
  • docs/cross-meeting-tools.md — MCP rollups across meeting summaries (list_action_items, digest)
  • docs/mcp-ui-recent-meetings.md — the MCP server's interactive recent-meetings UI surface
  • docs/auto-call-detection-spec.md — meeting auto-detection spec and phase status
  • docs/MEETING_CAPTURE_PROMPTING.md — draft spec for meeting-capture prompting
  • docs/ui-settings-menubar-spec.md — settings and menubar product intent
  • docs/DESIGN_TOKENS.md — source of truth for type, spacing, and corner-radius tokens
  • docs/release-guardrails.md — what may ship without an explicit publish decision
  • docs/release-notes-template.md — template for drafting release notes from main
  • docs/board-scorecard.md — the agent-runnable repo health task list
  • docs/posthog-100-wau-dashboard.md — minimum PostHog setup for steering toward 100+ WAU
  • docs/posthog-dashboard-query-helpers.md — the shared query catalog behind posthog-dashboard-queries.py
  • docs/posthog-product-intelligence-map.md — dated audit of which events answer which product question
  • docs/posthog-product-learning-plan.md — dated plan for what the analytics lane should learn next
  • Tests/README.md — verification surfaces and fast-test runner behavior
  • .agents/test-matrix.yml — quick path-to-verification map for agents
  • .agents/qa-gates.yml — product-risk-to-proof gate map for agents
  • Sources/*/CLAUDE.md — subsystem-local ownership and verification notes

Historical Zones

Treat these as reference, not current product surface:

  • archive/backend-beta-worker/

.claude/ is live tooling, not a historical zone: it holds the transcripted-qa skill and the humanize/tests/push slash commands.