# Daily Review > Daily Review (Meta-Skill) — orchestrates child skills: gather browser history, AI transcripts, app usage, process Telegram inbox, process Chrome bookmarks, snapshot portfolio DAU, sync all projects, suggest agent tasks, suggest today's plan, sync to calendar, and save the review. - Skill: `razbakov/daily-review` (Agent Skill) - Install (CLI): `npx skillmds@latest add razbakov/daily-review` - Raw SKILL.md: https://api.skillmd.com/api/skills/razbakov/daily-review/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: AI & ML - Author: razbakov (https://skillmd.com/u/razbakov) - Updated: 2026-09-21 - Page: https://skillmd.com/skills/razbakov/daily-review --- > **⚠️ Migration in progress (razbakov/ikigai#73):** Control Center has moved from Notion to GitHub Issues + Project v2. When this skill references creating/reading/updating Notion cards or pages, translate to GitHub equivalents: > - **Create card** → `gh issue create --repo --title "" --label agent:<name> --body "<S3 body>"` then `gh project item-add 5 --owner razbakov --url <issue-url>` > - **Read cards** → `gh issue list --repo <repo> --label agent:<name> --state open` (or across repos via `gh search issues "org:razbakov label:agent:<name> state:open"`) > - **Update card status** → move on board: `gh project item-edit` with the Status field, or close via `gh issue close` > - **Board columns**: Inbox → To do → In progress → To review → Done > - **Do not call any `notion-*` MCP tools** — the Notion MCP is disabled. # Daily Review (Meta-Skill) Run from the ikigai project directory. Gathers yesterday's activity, syncs all project next steps (with full git analysis), suggests today's plan, and adds approved plan to Google Calendar. Saves the review to `sessions/YYYY-MM-DD-daily-review.md`. This meta-skill orchestrates the following child skills in order: 1. **daily-review-browser-history** — Gather yesterday's Chrome browser history 2. **daily-review-ai-transcripts** — Gather yesterday's AI session transcripts 3. **daily-review-app-usage** — Query macOS app focus time 4. **process-inbox** — Collect Telegram + process Notion "To do" cards (GTD full pipeline) 5. **daily-review-browser-bookmarks** — Process Chrome bookmarks (GTD) 5.5. **analytics-check** — Snapshot portfolio DAU. Read by multiple agents daily: Viktor (tracking health, zero = bug), Luna (content/SEO levers), Kai (community/partnership levers for event-driven projects), project coordinators (own their KR). 6. **daily-review-sync-projects** — Discover, pull, analyze all projects 6.5. **suggest-tasks** — Propose agent tasks with assignments (Viktor/Luna/Marco/Kai) 7. **daily-review-suggest-plan** — Propose today's schedule with OKR view 8. **daily-review-calendar-sync** — Create Google Calendar events 9. **daily-review-save-review** — Save complete review to sessions file ## Trigger User says "daily review", "let's review projects", "daily check-in", or invokes `/daily-review`. ## Context - **Working directory:** `~/Projects/ikigai` - **Google account:** From CLAUDE.md Personal Info - **Calendar tool:** `gog cal` - **Projects directory:** `~/Projects/` - **Project inventory:** `PROJECTS.md` in ikigai workspace - **OKRs and focus:** `README.md` in ikigai workspace - **Session storage:** `sessions/` ## Execution ### Phase 1 — Gather Yesterday's Data (parallel) Run steps 1-5 in parallel where possible to save time. #### Step 1: Browser History Invoke: `//daily-review-browser-history` Captures: `sessions/YYYY-MM-DD-browser.md` #### Step 2: AI Transcripts Invoke: `//daily-review-ai-transcripts` Captures: `sessions/YYYY-MM-DD-ai-sessions.md` #### Step 3: App Usage Invoke: `//daily-review-app-usage` Depends on: Step 1 (appends to browser.md) Captures: App usage section in `sessions/YYYY-MM-DD-browser.md` #### Step 4: Process Inbox (Telegram + Notion) Invoke: `//process-inbox` Captures: Telegram export → Notion cards → agent dispatch. Saves `inbox/YYYY-MM-DD-HH-MM.md`, creates Control Center cards, dispatches agents for "To do" items. #### Step 5: Browser Bookmarks Invoke: `//daily-review-browser-bookmarks` Captures: `inbox/YYYY-MM-DD-HH-MM-bookmarks.md`, updated project docs #### Step 5.5: Portfolio DAU snapshot Invoke: `//analytics-check` (portfolio mode, no save) Captures: per-project DAU table in the daily review file. DAU is a shared organ — multiple agents read the snapshot daily to pull different levers: - **Viktor (CTO)** — tracking health. Any project at 0 events or unexpected silence on a project claimed to be live = SDK bug, fix within 48h. - **Luna (Content & Growth)** — content/SEO levers. Below-target DAU on a content-driven project = "what gets published or shared today to move it?" - **Kai (Community & Partnerships)** — community/partnership levers for event-driven projects. Below-target DAU on event projects = "which Munich channel, IG account, or partner can route traffic today?" - **Project coordinators** — own their project's KR; pull every lever available. Keep the daily output short: one row per project with yesterday's DAU, a delta vs the 7-day baseline, and a flag column noting which agent should look at it (tracking gap → Viktor; flat above 0 → Luna or Kai based on project type; hitting target → nobody). ### Phase 2 — Sync & Plan (sequential) #### Step 6: Sync Projects Invoke: `//daily-review-sync-projects` Depends on: Steps 4-5 (Telegram and bookmark items may add tasks to projects) Captures: Numbered project list with a/b/c task IDs, updated PROJECTS.md **User interaction:** "Which projects do you want to focus on today?" #### Step 6.5: Suggest Agent Tasks Invoke: `//suggest-tasks` Depends on: Step 6 (needs project list, PROJECTS.md, and Notion state) Captures: "Suggested" cards in Notion Control Center with agent assignments (Viktor/Luna/Marco/Kai) No user interaction needed — cards are created silently. The suggest-plan step (7) will mention how many agent tasks were suggested. #### Step 7: Suggest Plan Invoke: `//daily-review-suggest-plan` Depends on: Steps 1-6 (needs yesterday's data + project list with task IDs) Captures: OKR view with proposed schedule **User interaction:** "Does this plan work, or do you want to adjust?" #### Step 8: Calendar Sync Invoke: `//daily-review-calendar-sync` Depends on: Step 7 (needs approved schedule) Captures: List of calendar events created #### Step 9: Save Review Invoke: `//daily-review-save-review` Depends on: All previous steps Captures: `sessions/YYYY-MM-DD-daily-review.md` **User interaction:** Personal journaling prompt. (Scrum standup is handled separately by `/daily-standup` at 9am.) ## Notes - Each child skill can also be run independently via `/daily-review-{name}` - If a step fails, fix the issue and re-run that specific child skill - If the user says "skip yesterday" — jump straight to Step 6 - Steps 1, 2, 4, 5, and 5.5 have no dependencies on each other and should run in parallel - Step 3 depends on Step 1 (appends to the same file) - Sub-agents in Step 6 should run in parallel for speed — one agent per project - If today is Saturday, Step 7 will remind about the weekly review - **Always print the full output (project list + OKR view) to chat.** Saving to file is not a substitute for showing the user. This applies to BOTH interactive and automated/scheduled runs — the user reads the output after the fact - Actionable items from Telegram inbox and bookmarks should be merged into the project list - **Automated runs:** When running as a scheduled task without the user present, still execute ALL steps including full project sync (Step 6). Skip only steps that truly require real-time user input (plan approval, journaling). Always produce the complete report with git pull table + numbered project list + OKR view - **Project paths:** Many repos are nested. Always read the Project Path Registry from CLAUDE.md — never assume `~/Projects/<name>/.git`