# Year > Generates a strategic yearly review in the yearly note: year in numbers with consistency and peak/dead months, intellectual diet breakdown by source type and category, multi-quarter theme trajectory (emergent vs abandoned vs core), project portfolio with completion rate and multi-year drag, synthesis ratio trend across quarters, year in one sentence, and one data-derived priority for next year. Creates the yearly note if it doesn't exist. INVOKE when the user wants a yearly review or year summary. Triggers: 'yearly review', 'year review', 'year summary', 'итоги года', 'годовой обзор', 'обзор года', 'что за год'. - Skill: `flowing-abyss/year` (Agent Skill) - Install (CLI): `npx skillmds@latest add flowing-abyss/year` - Raw SKILL.md: https://api.skillmd.com/api/skills/flowing-abyss/year/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Coding & Dev Tools - Author: flowing-abyss (https://skillmd.com/u/flowing-abyss) - Updated: 2026-09-10 - Page: https://skillmd.com/skills/flowing-abyss/year --- # Yearly Review Generates a comprehensive yearly review without asking questions. Chains data upward from quarterly and monthly reviews, recomputes where those don't exist, and writes a `[!yearly-review]-` block into the yearly note, then appends a minimal space for the user's own reflections. ## Language Detection Before generating any content, detect the user's writing language. Read the 5 most recent daily notes (`periodic/daily/YYYY-MM-DD.md`). In non-frontmatter, non-wikilink, non-tag text, count Cyrillic vs Latin characters. If Cyrillic > 30% → write the entire review in Russian. Otherwise → English. Apply to ALL section headers, metric labels, table column names, inference lines (→), characterization labels, and participation prompts. Callout types and wikilinks remain unchanged regardless of detected language. ## Yearly Note Format **Path:** `periodic/yearly/YYYY.md` **Frontmatter (exact from template):** ```yaml --- tags: - periodic/year reviewed: false cssclasses: - hide-backlinks created: YYYY-MM-DDTHH:mm:ssZ updated: YYYY-MM-DDTHH:mm:ssZ --- ``` Note: yearly notes have **no `up:` field** — they are the top of the hierarchy. If the yearly note exists — find `[!yearly-review]-` and replace it. Otherwise create with template, then append after `[!success]-`. ## Execution Flow ``` User: "yearly review" / "итоги года" / "годовой обзор" | 1. Determine current year (YYYY) | 2. Find or create periodic/yearly/YYYY.md | 3. Detect language from last 5 daily notes | 4. Gather IN PARALLEL: a. Year in numbers — velocity across all 12 months b. Intellectual diet — source breakdown by type and category c. Theme trajectory — Q1 vs Q4 themes, emergent/abandoned/core; also read user-written sections below [!quarterly-review]- in each quarterly note d. Project portfolio — all year outcomes + multi-year drag e. Synthesis ratio — per quarter trend f. Previous yearly note — for year-over-year comparison | 5. For each section: collect data → derive one "→" inference from the specific pattern | 6. Generate review: numbers → diet → trajectory → projects → synthesis → year sentence → next year priority | 7. Write [!yearly-review]- callout into yearly note | 8. Append user participation section below the callout (outside it) | 9. Minimal chat confirmation ``` ## Year in Numbers Compute velocity for each of the 12 months, then aggregate. **Per-month computation (all 12 months in parallel):** ```bash # Notes created (run for each MM from 01 to 12): obsidian search query="[created: YYYY-MM]" path=base/notes total obsidian search query="[created: YYYY-MM]" path=base/additions total # Tasks closed (run for each MM): obsidian search query="✅ YYYY-MM" total # Sources processed (run for each MM): obsidian search query="[status: 🟩] [end: YYYY-MM]" path=sources total ``` **Aggregate:** - Total for year - Peak month: month with highest combined activity - Dead months: months where notes=0 AND tasks=0 AND sources=0 - Consistency score: (12 - dead months) / 12 × 100% *Consistency = share of months with any activity. 100% = active every month. Lower = feast-or-famine pattern.* **Year-over-year:** If `periodic/yearly/YYYY-1.md` exists and has a `[!yearly-review]-` block, extract its totals for comparison. Show delta with ↑/↓. ```markdown > **Year in numbers:** > - Notes: 187 (↑ from 134 last year) > - Tasks closed: 412 (↓ from 489) > - Sources processed: 43 (↑ from 31) > - Peak month: October (notes: 28, tasks: 67, sources: 9) > - Dead months: 2 (February, July) > - Consistency: 83% (10 of 12 months active) ``` **→ Inference:** After showing numbers, derive one data-specific observation. Look for: - All metrics ↑ vs last year → growth year; what structural change made this possible? - One metric ↓ while others ↑ → something shifted (notes ↑ but tasks ↓ = more thinking, less doing) - Dead months > 3 → sporadic pattern; the habit is fragile — what collapsed in those months? - Consistency < 60% → the work is concentrated in bursts; useful to understand what sustains the active periods Write the inference in detected language. Skip if no clear pattern. ## Intellectual Diet Breakdown of all sources completed (`end: YYYY-` + `status: 🟩`) by type and category. ### Source types All valid source tags in this vault: ``` source/article/paper source/article/resource source/book source/course source/cinematic/movie source/cinematic/series source/cinematic/anime source/podcast source/video/recording source/video/playlist source/music/album source/music/tracklist source/game ``` ```bash # Get all sources completed this year, then count by tag: obsidian search query="[status: 🟩] [end: YYYY]" path=sources # From those files, count occurrences of each tag above. # Or look up each type directly: obsidian tag name="source/book" verbose obsidian tag name="source/article/paper" verbose obsidian tag name="source/article/resource" verbose obsidian tag name="source/course" verbose obsidian tag name="source/cinematic/movie" verbose obsidian tag name="source/cinematic/series" verbose obsidian tag name="source/cinematic/anime" verbose obsidian tag name="source/podcast" verbose obsidian tag name="source/video/recording" verbose obsidian tag name="source/video/playlist" verbose obsidian tag name="source/music/album" verbose obsidian tag name="source/music/tracklist" verbose obsidian tag name="source/game" verbose # Cross-reference each result with the completed set for YYYY. ``` ### Source categories ```bash # Read completed sources from: obsidian search query="[status: 🟩] [end: YYYY]" path=sources # Extract category: frontmatter field from each file → count by category, show top 5 ``` ```markdown > **Intellectual diet:** > - By type: 📚 books: 12 | 📄 articles/resources: 11 | 📑 papers: 7 | 🎬 video/recordings: 5 | 📺 video/playlists: 3 | 🎓 courses: 2 | 🎬 movies: 2 | 📡 podcasts: 1 > - Top categories: [[knowledge management]] (11), [[artificial intelligence]] (9), [[productivity]] (7), [[writing]] (5), [[health]] (3) > - Dominant type: books (29% of total) ``` If one type > 60% of total — flag it: `⚠️ Heavy bias toward [type] — consider diversifying sources.` **→ Inference:** After showing diet, derive one data-specific observation. Look for: - Diet dominated by passive formats (articles, videos) vs active (books, courses, papers) → breadth vs depth signal - Top categories in diet diverge from core themes in knowledge base → are you studying what you're actually thinking about? - Very narrow category coverage (2–3 categories dominate) → potential echo chamber; what's being ignored? - Diet is varied and matches theme trajectory → strong alignment between learning and thinking Write the inference in detected language. Skip if no clear pattern. ## Multi-Quarter Theme Trajectory Reveals what emerged, what faded, and what held throughout the year. Also surfaces what the user themselves identified as important through their quarterly focus statements. ### Data source (priority order) 1. **Quarterly reviews** — if `periodic/quarterly/YYYY-Q1.md` through `YYYY-Q4.md` exist and have `[!quarterly-review]-` blocks, extract their "Theme Drift" sections 2. **User-written quarterly sections** — read text BELOW the `[!quarterly-review]-` callout in each quarterly note (user's "Focus for next quarter" and "My conclusions"). Surface any stated intentions that connect to theme patterns. 3. **Fallback** — grep wikilinks in daily notes for Q1 (`YYYY-01` to `YYYY-03`) and Q4 (`YYYY-10` to `YYYY-12`), count frequency per meta-note ### Classification - **Emergent:** appeared in Q3 or Q4, absent in Q1 and Q2 (new focus area this year) - **Abandoned:** prominent in Q1/Q2, absent in Q3/Q4 (lost momentum) - **Core:** present in all 4 quarters (sustained intellectual interest) - **Seasonal:** appears in exactly one quarter ```markdown > **Theme trajectory:** > - Emergent: [[machine learning]] (appeared Q3, dominant Q4), [[writing systems]] (Q4 only) > - Abandoned: [[linux administration]] (Q1 top, gone by Q3), [[economics]] (Q1–Q2 only) > - Core (all year): [[knowledge management]], [[learning]], [[productivity]] > - Seasonal: [[health]] (Q1 only) ``` **→ Inference:** After showing trajectory, derive one data-specific observation. Look for: - Many abandoned vs few core → intellectually nomadic year; broad exploration without depth - Strong emergent theme in Q4 → discovered something important late; worth making it explicit next year - Core theme with no completed projects → consistent interest but no shipping; what's the gap? - User's quarterly focus statements align with actual theme data → intentional direction held all year Write the inference in detected language. Skip if no clear pattern. ## Project Portfolio Full year view of project activity and execution rate. ### Completed this year ```bash obsidian search query="tag:#project [status: 🟩]" obsidian search query="tag:#project [status: 📢]" # Filter: keep only files where updated: frontmatter falls within YYYY ``` ### Frozen this year ```bash obsidian search query="tag:#project [status: ❄]" # Filter: keep only files where updated: frontmatter falls within YYYY ``` ### Started this year ```bash obsidian search query="tag:#project [status: 🟦]" # in progress obsidian search query="tag:#project [status: 🟥]" # todo # Filter: keep only files where created: frontmatter falls within YYYY ``` ### Multi-year drag ```bash obsidian search query="tag:#project [status: 🟦]" # in progress obsidian search query="tag:#project [status: 🟥]" # todo # Filter: keep only files where created: predates YYYY (from previous years) ``` **Completion rate** = completed / (completed + frozen + multi-year drag) × 100% *Of all projects that reached any resolution this year (completed or frozen) plus those carried from before, what share actually finished?* ```markdown > **Project portfolio:** > - Completed (5): [[Blog Series Part 1]] 📢, [[Home Lab Setup]] 🟩, [[Reading Tracker v2]] 🟩, [[Podcast Outline]] 📢, [[Note Sync Script]] 🟩 > - Frozen (3): [[Course Notes App]] ❄️, [[Spanish Learning Plan]] ❄️, [[Video Essay Draft]] ❄️ > - Started (8): ... > - Multi-year drag (2): [[Archive Reorganization]] 🟦 (from 2024), [[API Integration]] 🟥 (from 2023) > - Completion rate: 42% (5 completed / 12 resolvable) ``` If completion rate < 30%: `⚠️ Low completion rate — starting is easy, finishing is the bottleneck.` If multi-year drag > 2 projects: `⚠️ N projects carried from previous years — consider archiving or resetting.` **→ Inference:** After listing portfolio, derive one data-specific observation. Look for: - High frozen count → projects are being started optimistically but scope or motivation doesn't hold - Multi-year drag accumulating year over year → the system isn't clearing historical commitments - Started count >> Completed count → are projects being designed too large? Or is scope expanding after start? - Completion rate > 60% → strong execution year; what specifically worked that could be repeated? Write the inference in detected language. ## Synthesis Ratio Trend Shows whether synthesis efficiency improved or declined across the year. Compute ratio (notes created / sources processed) for each quarter: ```markdown > **Synthesis ratio:** > > | Q1 | Q2 | Q3 | Q4 | Year | > |-----|-----|-----|-----|------| > | 1.4 | 2.1 | 3.2 | 1.8 | 2.1 | > > *Scale: >2.0 = strong synthesis, 1.0–2.0 = balanced, 0.5–1.0 = consuming > creating, <0.5 = synthesis gap* > Peak: Q3. Trend: ↑ Q1→Q3, ↓ Q3→Q4. > Year average: 2.1 (balanced). ``` **→ Inference:** After showing trend, derive one data-specific observation. Look for: - Synthesis dropping consistently Q1→Q4 → reading appetite grew faster than processing capacity - Synthesis high in low-activity months → when less is happening, more gets processed - Consistently low (<1.0) all year → systematic habit gap; each new source needs a note before the next one opens - Peak synthesis quarter matches high project completion quarter → synthesis and shipping are connected Write the inference in detected language. ## The Year in One Sentence A single factual characterization derived from the dominant pattern — not inspirational, not motivational. Just accurate. Derive from the data: | Dominant pattern | Sentence type | |---|---| | Synthesis ratio < 1 across most quarters | "Год потребления: больше читал, чем создавал" | | Completion rate < 30% | "Год незавершённого: много стартов, мало финишей" | | One theme > 40% of all mentions | "Год одной темы: всё вращалось вокруг [X]" | | Consistency < 50% | "Год рывков: продуктивные месяцы чередовались с провалами" | | Many emergent themes, few core | "Год разброса: много новых направлений без фокуса" | | Completion rate > 60%, synthesis > 2 | "Год зрелости: высокий выхлоп и завершение начатого" | | Velocity ↑ YoY and synthesis ↑ | "Год роста: по всем метрикам лучше, чем предыдущий" | ```markdown > **Year in one sentence:** > Год одной темой: knowledge management занял 38% всех упоминаний — это либо глубина, либо туннельное зрение. ``` Write the label and characterization in detected language. One sentence. Name the exact metric or number that triggered the characterization. Do not soften or inflate — be accurate. ## Priority for Next Year One concrete direction derived from the year's weakest or most significant signal. Not a wish — a logical next step from the data. Also check user-written quarterly sections for any stated intentions about next year. Priority logic (pick the highest-priority applicable rule): 1. **Completion rate < 30%** → "Не начинать новые проекты пока не закроешь [N stuck projects]" 2. **Multi-year drag > 2** → "Первым делом: решить судьбу [oldest stuck project]" 3. **Synthesis ratio < 1 for 3+ quarters** → "Добавить шаг синтеза между чтением и следующим источником" 4. **Dead months > 3** → "Снизить объём в пиковые месяцы, чтобы не выгорать в мёртвые" 5. **Intellectual diet heavily biased (>60% one type)** → "Расширить диету: добавить [underrepresented source type]" 6. **Theme abandoned mid-year with no notes** → "[[Abandoned theme]] требует либо закрытия, либо осознанного возврата" 7. **Emergent theme dominant in Q4** → "[[Emergent theme]] оказался неожиданно важным — стоит сделать его явным приоритетом" ```markdown > **Priority for next year:** > Коэффициент синтеза падал 3 квартала подряд — добавить промежуточный шаг между обработкой источника и следующим чтением. Например: заметка до открытия нового источника. ``` One actionable sentence in detected language, with the specific number or fact that justifies it. ## Output Format ```markdown > [!yearly-review]- Yearly review — YYYY > **Year in numbers:** > - Notes: N (↑/↓ from last year) > - Tasks closed: N > - Sources processed: N > - Peak month: Month (notes: N, tasks: N, sources: N) > - Dead months: N (Month, Month) > - Consistency: N% (N of 12 months active) > → [inference about year pattern — only if clear] > > **Intellectual diet:** > - By type: 📚 books: N | 📄 articles: N | 🎬 video: N | ... > - Top categories: [[cat]] (N), [[cat]] (N) > → [inference about diet composition — only if clear] > > **Theme trajectory:** > - Emergent: [[theme]] (appeared QN) > - Abandoned: [[theme]] (gone by QN) > - Core: [[theme]], [[theme]] > → [inference about intellectual arc of the year — only if clear] > > **Project portfolio:** > - Completed (N): [[Project]] 📢, [[Project]] 🟩 > - Frozen (N): [[Project]] ❄️ > - Started (N): N projects > - Multi-year drag (N): [[Project]] (from YYYY) > - Completion rate: N% > → [inference about execution capacity — only if clear] > > **Synthesis ratio:** > > | Q1 | Q2 | Q3 | Q4 | Year | > |----|----|----|----|----| > | N | N | N | N | N | > > *Scale: >2.0 = strong, 1.0–2.0 = balanced, 0.5–1.0 = consuming, <0.5 = gap* > → [inference about synthesis trend — only if clear] > > **Year in one sentence:** > [Characterization label]: [one sentence in detected language] > > **Priority for next year:** > [One actionable sentence] ``` **Rules:** - Write into note — never just output to chat - Collapsed by default (`-` suffix) - Callout title includes year: `[!yearly-review]- Yearly review — 2026` - All labels translated to detected language - Skip sections with no content - Skip "→" inference lines where no clear pattern exists — never write generic lines - Every reference = `[[wikilink]]` - Flags (`⚠️`) only for patterns that cross clear thresholds — don't flag minor deviations ## User Participation Sections After the `[!yearly-review]-` callout, appended **outside it** (in the note body), with a horizontal rule separator. The yearly note is the top of the hierarchy — there is no higher-level review to feed. This section is for the user's personal synthesis: what the year actually meant, what they want to carry forward. **Default form (always):** ```markdown --- **Мои выводы за год / My reflections on the year:** **Намерение на следующий год / My intention for [YYYY+1]:** ``` **Extended form (when review data shows a specific strong signal):** ```markdown --- *[One direct question from the review's most significant finding — e.g., "[[X theme]] оказался заброшен к Q3 без единой заметки — это завершено или просто забыто?" / "[[X theme]] was abandoned by Q3 without a single note — resolved or forgotten?"]* **Мои выводы за год / My reflections on the year:** **Намерение на следующий год / My intention for [YYYY+1]:** ``` Write the italic question in detected language. Use extended form only when the signal is unambiguous. Unlike lower-level review notes, this is not read by a higher-level review. This section exists purely for the user — a space for personal synthesis at the highest level of the review system. ## Chat Output ``` Review written → periodic/yearly/2026.md Notes: 187 (↑ from 134). Tasks: 412. Sources: 43. Diet: articles dominant (42%). Top: knowledge management, AI. Themes: knowledge management core all year; linux abandoned Q3. Projects: 5/12 completed (42%). 2 multi-year drag. Synthesis: avg 2.1, peak Q3. Year: [characterization in 1 line]. Priority: [priority in 1 line]. ``` If note was created — first line: `Created periodic/yearly/2026.md + review.` Adapt language of confirmation to match detected language.