{"id":"88e45866-322f-4ea1-9f7c-f4255e067b3c","shortId":"GLAr2J","kind":"skill","title":"weekly-ai-workflow-review","tagline":"Analyze weekly Claude interaction notes to reveal delegation patterns, effective prompts, and optimization areas. Use for weekly AI workflow review, reflecting on Claude usage, improving prompts. Triggers: 'weekly ai workflow review', 'review my claude interactions', 'еженедельны","description":"# Weekly AI Workflow Review\n\nThis skill analyzes weekly notes about tasks delegated to Claude and produces a structured reflection report: what worked, what needed rework, recurring task types, and recommended prompt templates for the next week.\n\n**Input:**\n- Markdown file or pasted text with weekly Claude task log (task descriptions, prompts or their summaries, outcomes: used / edited / discarded)\n\n**Output:**\n- Structured markdown reflection report (inline in chat or saved as `weekly-ai-review-YYYY-MM-DD.md` on request)\n\n---\n\n## Language Detection\n\nDetect the user's language from their message:\n- If Russian (or contains Cyrillic): respond in Russian\n- If English (or other Latin-script language): respond in English\n- If ambiguous: respond in the language of the trigger phrase used\n\n**Localization note:** When responding in Russian, translate all section headers and prompts in the Output Format template:\n- \"Weekly AI Workflow Review\" → \"Еженедельный обзор AI-рабочего процесса\"\n- \"Period\" → \"Период\", \"Tasks analyzed\" → \"Задач проанализировано\"\n- \"Patterns of the Week\" → \"Паттерны недели\"\n- \"What Worked Well\" → \"Что работало хорошо\"\n- \"Areas for Improvement\" → \"Зоны улучшения\"\n- \"Recommended Prompt Templates\" → \"Рекомендованные промпт-шаблоны\"\n- \"Takeaway\" → \"Итог\"\n\n---\n\n## Instructions\n\n### Step 1: Validate and Parse Input\n\n1. Check that input contains at least 1 task description or prompt\n   - If input is empty or contains only a file header: stop and return: \"Task log not found. Please paste your weekly list of tasks delegated to Claude.\"\n   - If input appears unrelated to AI/Claude interactions (e.g., general journal, shopping list): return: \"No Claude-related tasks found in the provided notes. Please include descriptions of tasks you delegated to Claude this week.\"\n\n2. Parse entries from the input\n   - Accept both structured (markdown list) and unstructured (plain text) formats\n   - Extract for each entry: task description, prompt or summary, outcome if noted (used / edited / discarded / unclear)\n   - If entries are unstructured plain text: attempt to identify task boundaries by paragraph, line break, or numbered list patterns\n   - If task boundaries cannot be clearly identified (continuous spaghetti text): request a minimal structure: \"Unable to parse task boundaries from the text. Please provide tasks as a list (bullet points, numbers, or line breaks between entries) so I can analyze them accurately.\"\n\n3. Note sample size\n   - If fewer than 3 tasks identified: continue but add a note in the report: \"Small sample (N tasks) — patterns are indicative; consider logging at least 5 tasks per week for reliable trends.\"\n\n### Step 2: Classify Tasks by Type\n\n1. Group parsed entries into task types:\n   - **Content / Writing** — posts, emails, documents, summaries\n   - **Analysis** — data review, comparison, synthesis\n   - **Research** — topic exploration, fact-finding\n   - **Formatting / Editing** — structure, grammar, style\n   - **Ideation** — brainstorming, generation of options\n   - **Other** — anything that doesn't fit above categories\n\n2. Count tasks per type; identify the dominant type(s)\n\n3. Note recurring tasks — any task type or topic that appears 2+ times\n\n**Edge Cases:**\n- If no outcome is noted for any task: skip outcome-based analysis; analyze task types and prompt patterns only; mark report section as \"Outcome data unavailable\"\n- If all tasks belong to a single type: note in report, skip cross-type comparison\n\n### Step 3: Identify Successful Interactions\n\n1. Mark as **successful** any task where:\n   - Outcome is explicitly noted as \"used\" or \"no edits\"\n   - Or: description implies result was accepted (e.g., \"sent it\", \"published\", \"approved\")\n\n2. Extract the prompt pattern for each successful interaction (verb + object + context)\n\n3. Note what made these prompts effective: specificity, role assignment, format instruction, example provided\n\n### Step 4: Identify Pain Points\n\n1. Mark as **problematic** any task where:\n   - Outcome is noted as \"edited significantly\", \"multiple iterations\", \"discarded\", or \"not used\"\n   - Or: description implies significant rework (e.g., \"rewrote half of it\", \"had to redo\")\n\n2. For each problematic task: diagnose the likely cause from the prompt/description:\n   - Vague request (no format, audience, or length specified)\n   - Missing context (no background or constraints)\n   - Scope too broad (asked for too much at once)\n   - Wrong tone or style (no style guidance provided)\n\n3. Generate a specific improvement recommendation for each pain point\n\n### Step 5: Generate Prompt Templates\n\n1. For each recurring task type (2+ occurrences): draft a reusable prompt template\n   - Template structure: `[Action] + [Object] + [Format/length] + [Audience/tone] + [Constraints]`\n   - Fill from successful interactions or improve from pain points\n   - Keep templates general enough to reuse, specific enough to be actionable\n\n2. Prioritize templates for task types that had both recurrence and pain points\n\n### Step 6: Compose Report\n\n1. Build the structured report using Output Format below\n2. Populate all sections; if a section has no data, write \"None found\" — do not omit the section\n3. If user asks to save: write output to `weekly-ai-review-YYYY-MM-DD.md` using today's date\n\n---\n\n## Output Format\n\n```markdown\n## Weekly AI Workflow Review\n**Period:** [week or date range if provided]  **Tasks analyzed:** N\n\n---\n\n### Patterns of the Week\n- Dominant task type: [type] (N tasks, X%)\n- Successful interactions: N (X%)\n- Tasks requiring rework: N (X%)\n- Recurring topics: [list or \"None\"]\n\n---\n\n### What Worked Well\n- **[Task type]:** [Description of successful interaction]. Prompt pattern: \"[excerpt or reconstruction]\"\n- **[Task type]:** [Second example if available]\n\n---\n\n### Areas for Improvement\n- **[Task]:** Required [N iterations / significant edits]. Likely cause: [diagnosis]. Recommendation: [specific prompt fix]\n- **[Task]:** [Second example if available]\n\n---\n\n### Recommended Prompt Templates\n1. For [task type]: \"[Template: action + object + format + constraints]\"\n2. For [task type]: \"[Template]\"\n\n---\n\n### Takeaway\n[1–2 sentences summarizing the week's AI workflow: what to keep, what to change]\n```\n\n**Field rules:**\n- Prompt templates must be specific and actionable (not \"be more specific\" but an actual template)\n- Diagnoses must be concrete (one of: vague request / missing context / scope too broad / wrong tone)\n- Takeaway is 1–2 sentences only; no bullet points\n\n---\n\n## Negative Cases\n\n- If input is empty or header-only: stop with \"Task log not found. Please paste your weekly list of tasks delegated to Claude.\"\n- If input contains no AI/Claude-related content: stop with \"No Claude-related tasks found. Please include descriptions of tasks you delegated to Claude this week.\"\n- If prompt patterns cannot be inferred from descriptions alone: note in report \"Prompt patterns could not be inferred — consider logging actual prompts for richer analysis.\"","tags":["weekly","workflow","review","claude","skills","kit","kirkruglov","agent-skills","agentic-skills","ai-agents","ai-skills","awesome-list"],"capabilities":["skill","source-kirkruglov","skill-weekly-ai-workflow-review","topic-agent-skills","topic-agentic-skills","topic-ai-agents","topic-ai-skills","topic-awesome-list","topic-claude","topic-claude-ai","topic-claude-ai-skills","topic-claude-code","topic-claude-cowork","topic-claude-memory","topic-claude-skills"],"categories":["claude-skills-kit"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/KirKruglov/claude-skills-kit/weekly-ai-workflow-review","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add KirKruglov/claude-skills-kit","source_repo":"https://github.com/KirKruglov/claude-skills-kit","install_from":"skills.sh"}},"qualityScore":"0.453","qualityRationale":"deterministic score 0.45 from registry signals: · indexed on github topic:agent-skills · 7 github stars · SKILL.md body (6,991 chars)","verified":false,"liveness":"unknown","lastLivenessCheck":null,"agentReviews":{"count":0,"score_avg":null,"cost_usd_avg":null,"success_rate":null,"latency_p50_ms":null,"narrative_summary":null,"summary_updated_at":null},"enrichmentModel":"deterministic:skill-github:v1","enrichmentVersion":1,"enrichedAt":"2026-05-18T19:13:39.590Z","embedding":null,"createdAt":"2026-05-18T13:21:17.667Z","updatedAt":"2026-05-18T19:13:39.590Z","lastSeenAt":"2026-05-18T19:13:39.590Z","tsv":"'1':214,219,226,427,542,600,690,747,874,889,938 '2':292,422,469,490,569,632,696,730,756,883,890,939 '3':385,392,479,538,581,675,774 '4':596 '5':414,686 '6':744 'accept':298,563 'accur':384 'action':705,729,879,912 'actual':919,1016 'add':397 'ai':3,23,34,43,171,177,792,896 'ai-рабочего':176 'ai/claude':263 'ai/claude-related':975 'alon':1004 'ambigu':143 'analysi':440,506,1020 'analyz':6,48,183,382,507,803 'anyth':462 'appear':260,489 'approv':568 'area':19,198,850 'ask':661,777 'assign':590 'attempt':330 'audienc':648 'audience/tone':708 'avail':849,870 'background':655 'base':505 'belong':524 'boundari':334,345,361 'brainstorm':457 'break':338,376 'broad':660,933 'build':748 'bullet':371,943 'cannot':346,999 'case':493,946 'categori':468 'caus':640,860 'chang':903 'chat':106 'check':220 'classifi':423 'claud':8,28,39,55,86,257,273,289,970,981,993 'claude-rel':272,980 'clear':348 'comparison':443,536 'compos':745 'concret':924 'consid':410,1014 'constraint':657,709,882 'contain':126,223,236,973 'content':434,976 'context':580,653,930 'continu':350,395 'could':1010 'count':470 'cross':534 'cross-typ':533 'cyril':127 'data':441,519,765 'date':787,798 'deleg':13,53,255,287,968,991 'descript':90,228,283,313,559,620,835,987,1003 'detect':114,115 'diagnos':637,921 'diagnosi':861 'discard':98,322,615 'document':438 'doesn':464 'domin':476,809 'draft':698 'e.g':265,564,624 'edg':492 'edit':97,321,452,557,611,858 'effect':15,587 'email':437 'empti':234,950 'english':132,141 'enough':722,726 'entri':294,311,325,378,430 'exampl':593,847,868 'excerpt':841 'explicit':551 'explor':447 'extract':308,570 'fact':449 'fact-find':448 'fewer':390 'field':904 'file':80,239 'fill':710 'find':450 'fit':466 'fix':865 'format':168,307,451,591,647,754,789,881 'format/length':707 'found':247,276,768,960,984 'general':266,721 'generat':458,676,687 'grammar':454 'group':428 'guidanc':673 'half':626 'header':162,240,953 'header-on':952 'ideat':456 'identifi':332,349,394,474,539,597 'impli':560,621 'improv':30,200,679,715,852 'includ':282,986 'indic':409 'infer':1001,1013 'inlin':104 'input':78,218,222,232,259,297,948,972 'instruct':212,592 'interact':9,40,264,541,577,713,817,838 'iter':614,856 'journal':267 'keep':719,900 'languag':113,119,138,147 'latin':136 'latin-script':135 'least':225,413 'length':650 'like':639,859 'line':337,375 'list':252,269,302,341,370,827,965 'local':153 'log':88,245,411,958,1015 'made':584 'mark':514,543,601 'markdown':79,101,301,790 'messag':122 'minim':355 'miss':652,929 'much':664 'multipl':613 'must':908,922 'n':405,804,813,818,823,855 'need':65 'negat':945 'next':76 'none':767,829 'note':10,50,154,280,319,386,399,480,498,529,552,582,609,1005 'number':340,373 'object':579,706,880 'occurr':697 'omit':771 'one':925 'optim':18 'option':460 'outcom':95,317,496,504,518,549,607 'outcome-bas':503 'output':99,167,753,781,788 'pain':598,683,717,741 'paragraph':336 'pars':217,293,359,429 'past':82,249,962 'pattern':14,186,342,407,512,573,805,840,998,1009 'per':416,472 'period':180,795 'phrase':151 'plain':305,328 'pleas':248,281,365,961,985 'point':372,599,684,718,742,944 'popul':757 'post':436 'priorit':731 'problemat':603,635 'produc':57 'prompt':16,31,72,91,164,204,230,314,511,572,586,688,701,839,864,872,906,997,1008,1017 'prompt/description':643 'provid':279,366,594,674,801 'publish':567 'rang':799 'recommend':71,203,680,862,871 'reconstruct':843 'recur':67,481,693,825 'recurr':739 'redo':631 'reflect':26,60,102 'relat':274,982 'reliabl':419 'report':61,103,402,515,531,746,751,1007 'request':112,353,645,928 'requir':821,854 'research':445 'respond':128,139,144,156 'result':561 'return':243,270 'reus':724 'reusabl':700 'reveal':12 'review':5,25,36,37,45,173,442,794 'rework':66,623,822 'rewrot':625 'richer':1019 'role':589 'rule':905 'russian':124,130,158 'sampl':387,404 'save':108,779 'scope':658,931 'script':137 'second':846,867 'section':161,516,759,762,773 'sent':565 'sentenc':891,940 'shop':268 'signific':612,622,857 'singl':527 'size':388 'skill':47 'skill-weekly-ai-workflow-review' 'skip':502,532 'small':403 'source-kirkruglov' 'spaghetti':351 'specif':588,678,725,863,910,916 'specifi':651 'step':213,421,537,595,685,743 'stop':241,955,977 'structur':59,100,300,356,453,704,750 'style':455,670,672 'success':540,545,576,712,816,837 'summar':892 'summari':94,316,439 'synthesi':444 'takeaway':210,888,936 'task':52,68,87,89,182,227,244,254,275,285,312,333,344,360,367,393,406,415,424,432,471,482,484,501,508,523,547,605,636,694,734,802,810,814,820,833,844,853,866,876,885,957,967,983,989 'templat':73,169,205,689,702,703,720,732,873,878,887,907,920 'text':83,306,329,352,364 'time':491 'today':785 'tone':668,935 'topic':446,487,826 'topic-agent-skills' 'topic-agentic-skills' 'topic-ai-agents' 'topic-ai-skills' 'topic-awesome-list' 'topic-claude' 'topic-claude-ai' 'topic-claude-ai-skills' 'topic-claude-code' 'topic-claude-cowork' 'topic-claude-memory' 'topic-claude-skills' 'translat':159 'trend':420 'trigger':32,150 'type':69,426,433,473,477,485,509,528,535,695,735,811,812,834,845,877,886 'unabl':357 'unavail':520 'unclear':323 'unrel':261 'unstructur':304,327 'usag':29 'use':20,96,152,320,554,618,752,784 'user':117,776 'vagu':644,927 'valid':215 'verb':578 'week':2,7,22,33,42,49,77,85,170,189,251,291,417,791,796,808,894,964,995 'weekly-ai-review-yyyy-mm-dd.md':110,783 'weekly-ai-workflow-review':1 'well':194,832 'work':63,193,831 'workflow':4,24,35,44,172,793,897 'write':435,766,780 'wrong':667,934 'x':815,819,824 'еженедельны':41 'еженедельный':174 'задач':184 'зоны':201 'итог':211 'недели':191 'обзор':175 'паттерны':190 'период':181 'проанализировано':185 'промпт':208 'промпт-шаблоны':207 'процесса':179 'работало':196 'рабочего':178 'рекомендованные':206 'улучшения':202 'хорошо':197 'что':195 'шаблоны':209","prices":[{"id":"a6a4ac80-b737-4b81-98e5-5db140ad3696","listingId":"88e45866-322f-4ea1-9f7c-f4255e067b3c","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"KirKruglov","category":"claude-skills-kit","install_from":"skills.sh"},"createdAt":"2026-05-18T13:21:17.667Z"}],"sources":[{"listingId":"88e45866-322f-4ea1-9f7c-f4255e067b3c","source":"github","sourceId":"KirKruglov/claude-skills-kit/weekly-ai-workflow-review","sourceUrl":"https://github.com/KirKruglov/claude-skills-kit/tree/main/skills/weekly-ai-workflow-review","isPrimary":false,"firstSeenAt":"2026-05-18T13:21:17.667Z","lastSeenAt":"2026-05-18T19:13:39.590Z"}],"details":{"listingId":"88e45866-322f-4ea1-9f7c-f4255e067b3c","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"KirKruglov","slug":"weekly-ai-workflow-review","github":{"repo":"KirKruglov/claude-skills-kit","stars":7,"topics":["agent-skills","agentic-skills","ai","ai-agents","ai-skills","awesome-list","claude","claude-ai","claude-ai-skills","claude-code","claude-cowork","claude-memory","claude-skills","memory-management","productivity","productivity-tools","project-management"],"license":"mit","html_url":"https://github.com/KirKruglov/claude-skills-kit","pushed_at":"2026-05-18T04:27:46Z","description":"40+ curated agent skills for Claude Cowork and Claude.ai — ready-to-use tools for non-technical users: project management, productivity, and AI workflow automation","skill_md_sha":"85f52686930ee63165280a0ef6dce8425b6d00e3","skill_md_path":"skills/weekly-ai-workflow-review/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/KirKruglov/claude-skills-kit/tree/main/skills/weekly-ai-workflow-review"},"layout":"multi","source":"github","category":"claude-skills-kit","frontmatter":{"name":"weekly-ai-workflow-review","description":"Analyze weekly Claude interaction notes to reveal delegation patterns, effective prompts, and optimization areas. Use for weekly AI workflow review, reflecting on Claude usage, improving prompts. Triggers: 'weekly ai workflow review', 'review my claude interactions', 'еженедельный обзор AI-задач', 'паттерны работы с Claude'."},"skills_sh_url":"https://skills.sh/KirKruglov/claude-skills-kit/weekly-ai-workflow-review"},"updatedAt":"2026-05-18T19:13:39.590Z"}}