{"id":"c0b2e458-7c96-43fc-bfed-08bf0b2eec11","shortId":"YRJxUf","kind":"skill","title":"skill-optimizer-lawvable","tagline":"Guide to analyze a current work session and propose improvements to skills. Use (1) automatically after working with a skill to capture learnings, (2) when the user suggests improvements, corrections, or additions during a skill-related session, or (3) when the user manually invo","description":"# Self-Improve Skill\n\nAnalyze the current conversation and propose improvements to skills based on corrections, successes, and edge cases discovered during the work session.\n\n## Triggers\n\n- `self-improve` - Analyze session and propose improvements\n- `self-improve [skill-name]` - Target a specific skill\n- `self-improve on` - Enable automatic mode (hook)\n- `self-improve off` - Disable automatic mode\n- `self-improve status` - Show automatic mode status\n- `self-improve [skill-name] history` - Show modification history\n\n---\n\n## Main Workflow (`self-improve`)\n\n### Step 1: Identify the Skill\n\nIf skill name not provided, list available skills from `skills/` directory and ask:\n\n```\nWhich skill should I analyze for this session?\n[List skills found in skills/ directory]\n```\n\n### Step 2: Detect Signals\n\nScan the conversation for **signals** - moments where the user expressed feedback:\n\n| Signal Type | Examples |\n|-------------|----------|\n| **Correction** | \"No\", \"That's not right\", \"It's missing X\", \"Always do Y\", user rewrites output |\n| **Success** | \"Perfect\", \"Yes\", \"Exactly\", user accepts without changes |\n| **Edge case** | User needed a workaround, skill couldn't handle the request |\n\n### Step 3: Evaluate Each Signal for Quality\n\nFor each correction signal, evaluate if it can become a good skill instruction.\n\n#### Quality Criteria\n\n**1. COMPLETE**\n\nThe instruction includes all information needed to apply it. No need to look elsewhere or make assumptions.\n\n| Grade | Example |\n|-------|---------|\n| Pass | \"Structure output as: Key Terms / Risk Areas / Suggested Revisions\" |\n| Fail | \"Use the standard format\" (which format?) |\n| Fail | \"Follow our firm's guidelines\" (what guidelines?) |\n\n**2. PRECISE**\n\nNo vague or subjective terms. Two different people reading the instruction would understand it the same way.\n\n| Grade | Example |\n|-------|---------|\n| Pass | \"Flag non-compete clauses over 12 months as high risk\" |\n| Fail | \"Be more thorough in the analysis\" |\n| Fail | \"Make it more appropriate for clients\" |\n\n**3. ATOMIC**\n\nOne instruction addresses one single requirement. Multiple checks should be split into separate instructions.\n\n| Grade | Example |\n|-------|---------|\n| Pass | \"Check for governing law clause\" |\n| Fail | \"Check for governing law, jurisdiction, and arbitration clauses\" (three checks - split them) |\n\n**4. STABLE**\n\nIf referencing regulations or standards, specify the version or date. The instruction should be evaluable the same way regardless of when it's read.\n\n| Grade | Example |\n|-------|---------|\n| Pass | \"Review the termination provisions under our internal policy [policy name and reference], dated December 12, 2024.\" |\n| Fail | \"Follow latest market standards\" (which standards? will change over time) |\n\n### Step 4: Grade the Signal\n\n| Criteria Met | Action |\n|--------------|--------|\n| **All 4 criteria pass** | Add to skill directly |\n| **Less than 4 criteria** | Ask for clarification (see Step 5) |\n\n### Step 5: Ask for Clarification\n\nWhen feedback doesn't meet all criteria, ask for what's missing using the `AskUserQuestion` tool:\n\n```\nI detected a correction but need more information to improve the skill.\n\nYou said: \"[user's feedback]\"\n\nTo create a clearer instruction, I need the following information: \n\n[Structured tool call listing what's missing based on failed criteria]\n```\n\n**If the user provides clarification** → Update the instruction and proceed to Step 6.\n\n**If the user prefers the original** → Proceed to Step 6 with the original instruction.\n\n### Step 6: Propose Changes\n\n```\n--- Learning: [skill-name] ---\n\nProposed additions:\n\n1. \"[exact instruction to add]\"\n   Source: \"[quote from conversation]\"\n\n2. \"[exact instruction to add]\"\n   Source: \"[quote from conversation]\"\n\n---\n\nApply these changes? [Y/n]\n```\n\n### Step 7: If Approved\n\n1. **Update SKILL.md**\n    - Read `skills/[skill-name]/SKILL.md`\n    - Add each instruction in the appropriate section\n    - Each instruction must be readable and applicable on its own\n\n2. **Update `skills/[skill-name]/CHANGELOG.md`** \n    - Create if doesn't exist\n    - Add new entry AT THE TOP:\n      ```markdown\n      ## [DATE (format: \"January 7, 2026\")]\n      [Description of changes in natural language, 1-3 sentences]\n      ```\n    - Entry rules:\n      - Most recent at top\n      - 1-3 sentences max\n      - Natural language\n      - No git references\n\n### Step 8: Save Observations\n\nFor signals that couldn't be processed, offer to save:\n\n```\nSave these observations for later review?\n- \"[signal 1]\" - Status: [why insufficient]\n- \"[signal 2]\" - Status: [why insufficient]\n```\n\nIf yes, append to `skills/[skill-name]/OBSERVATIONS.md`\n\n---\n\n## Secondary Commands\n\n### self-improve on\n\n1. Run:\n    ```bash\n    rm -f ./.disabled\n    ```\n2. Reply: \"Automatic mode enabled.\"\n\n### self-improve off\n\n1. Run:\n    ```bash\n    touch ./.disabled\n    ```\n2. Reply: \"Automatic mode disabled.\"\n\n### self-improve status\n\nCheck `.disabled` file existence and report.\n\n## self-improve [skill-name] history\n\n1. Display CHANGELOG.md content\n2. Ask: \"Would you like to revert to a previous version?\"\n3. If yes:\n    - update the appropriate sections in `skills/[skill-name]/SKILL.md`\n    - update `skills/[skill-name]/CHANGELOG.md` with a rollback note\n    \n---\n\n## Examples\n\n### Example 1: All criteria met\n\n**User said:** \"Always flag non-compete clauses over 12 months as high risk\"\n\n**Evaluation:**\n- Complete: Yes - instruction is fully specified\n- Precise: Yes - \"12 months\" and \"high risk\" are clear\n- Atomic: Yes - single check\n- Stable: Yes - no time dependency\n\n**Result:** Add directly\n\n### Example 2: Missing criteria\n\n**User said:** \"Flag any non-market-standard indemnification clause\"\n\n**Evaluation:**\n- Complete: No - \"non-market-standard\" is not defined\n- Precise: No - \"market standard\" is subjective and varies by deal type\n- Atomic: Yes - single check\n- Stable: No - market standards evolve over time\n\n**Action:** Ask for clarification using the `AskUserQuestion` tool:\n```\nI detected a correction but need more details.\n\nYou said: \"Flag any non-market-standard indemnification clause\"\n\nTo make this actionable, can you specify:\n- What makes an indemnification clause \"non-market-standard\"? (e.g., uncapped liability, coverage of indirect damages, no carve-outs for gross negligence)\n\nDo you want to provide more details, or should I add the instruction as you stated it?\n```\n\n**If user clarifies:** Update the instruction and add it.\n**If user prefers the original:** Add the instruction as stated.\n\n---\n\n## Important Notes\n\n- Never guess what the user meant - always ask if unclear\n- Never infer requirements from context - they must be explicit\n- One instruction = one check - split bundled feedback\n- Fewer good instructions is better than many vague ones\n- CHANGELOG.md is the user-facing record","tags":["skill","optimizer","lawvable","awesome","legal","skills","agent-skills","automation","law","legal-work","workflows"],"capabilities":["skill","source-lawvable","skill-skill-optimizer-lawvable","topic-agent-skills","topic-automation","topic-law","topic-legal-work","topic-workflows"],"categories":["awesome-legal-skills"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/lawvable/awesome-legal-skills/skill-optimizer-lawvable","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add lawvable/awesome-legal-skills","source_repo":"https://github.com/lawvable/awesome-legal-skills","install_from":"skills.sh"}},"qualityScore":"0.605","qualityRationale":"deterministic score 0.60 from registry signals: · indexed on github topic:agent-skills · 310 github stars · SKILL.md body (6,835 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":"2026-05-02T18:53:56.780Z","embedding":null,"createdAt":"2026-04-18T22:04:42.419Z","updatedAt":"2026-05-02T18:53:56.780Z","lastSeenAt":"2026-05-02T18:53:56.780Z","tsv":"'-3':631,640 '/.disabled':698,712 '/changelog.md':606,768 '/observations.md':686 '/skill.md':582,762 '1':18,133,240,548,574,630,639,669,693,708,735,775 '12':314,413,788,802 '2':28,165,286,557,600,674,699,713,739,822 '2024':414 '2026':623 '3':44,219,333,750 '4':370,427,435,444 '5':451,453 '6':523,533,539 '7':571,622 '8':649 'accept':203 'action':433,867,896 'add':438,552,561,583,612,819,933,947,954 'addit':36,547 'address':337 'alway':192,781,967 'analysi':325 'analyz':7,54,79,154 'append':680 'appli':249,566 'applic':596 'appropri':330,588,755 'approv':573 'arbitr':364 'area':268 'ask':149,446,454,464,740,868,968 'askuserquest':471,873 'assumpt':258 'atom':334,809,856 'automat':19,99,107,114,701,715 'avail':143 'base':63,507 'bash':695,710 'becom':233 'better':991 'bundl':985 'call':502 'captur':26 'carv':918 'carve-out':917 'case':69,207 'chang':205,423,541,568,626 'changelog.md':737,996 'check':342,352,358,367,722,812,859,983 'clarif':448,456,515,870 'clarifi':942 'claus':312,356,365,786,834,892,904 'clear':808 'clearer':493 'client':332 'command':688 'compet':311,785 'complet':241,794,836 'content':738 'context':975 'convers':57,170,556,565 'correct':34,65,182,227,476,878 'couldn':213,655 'coverag':912 'creat':491,607 'criteria':239,431,436,445,463,510,777,824 'current':9,56 'damag':915 'date':381,411,619 'deal':854 'decemb':412 'defin':844 'depend':817 'descript':624 'detail':882,929 'detect':166,474,876 'differ':294 'direct':441,820 'directori':147,163 'disabl':106,717,723 'discov':70 'display':736 'doesn':459,609 'e.g':909 'edg':68,206 'elsewher':255 'enabl':98,703 'entri':614,633 'evalu':220,229,386,793,835 'evolv':864 'exact':201,549,558 'exampl':181,260,306,350,397,773,774,821 'exist':611,725 'explicit':979 'express':177 'f':697 'face':1001 'fail':271,278,319,326,357,415,509 'feedback':178,458,489,986 'fewer':987 'file':724 'firm':281 'flag':308,782,827,885 'follow':279,416,498 'format':275,277,620 'found':160 'fulli':798 'git':646 'good':235,988 'govern':354,360 'grade':259,305,349,396,428 'gross':921 'guess':962 'guid':5 'guidelin':283,285 'handl':215 'high':317,791,805 'histori':123,126,734 'hook':101 'identifi':134 'import':959 'improv':14,33,52,60,78,83,86,96,104,111,119,131,482,691,706,720,730 'includ':244 'indemnif':833,891,903 'indirect':914 'infer':972 'inform':246,480,499 'instruct':237,243,298,336,348,383,494,518,537,550,559,585,591,796,935,945,956,981,989 'insuffici':672,677 'intern':405 'invo':49 'januari':621 'jurisdict':362 'key':265 'languag':629,644 'later':666 'latest':417 'law':355,361 'lawvabl':4 'learn':27,542 'less':442 'liabil':911 'like':743 'list':142,158,503 'look':254 'main':127 'make':257,327,894,901 'mani':993 'manual':48 'markdown':618 'market':418,831,840,847,862,889,907 'max':642 'meant':966 'meet':461 'met':432,778 'miss':190,468,506,823 'mode':100,108,115,702,716 'modif':125 'moment':173 'month':315,789,803 'multipl':341 'must':592,977 'name':89,122,139,408,545,581,605,685,733,761,767 'natur':628,643 'need':209,247,252,478,496,880 'neglig':922 'never':961,971 'new':613 'non':310,784,830,839,888,906 'non-compet':309,783 'non-market-standard':829,838,887,905 'note':772,960 'observ':651,664 'offer':659 'one':335,338,980,982,995 'optim':3 'origin':529,536,953 'out':919 'output':197,263 'pass':261,307,351,398,437 'peopl':295 'perfect':199 'polici':406,407 'precis':287,800,845 'prefer':527,951 'previous':748 'proceed':520,530 'process':658 'propos':13,59,82,540,546 'provid':141,514,927 'provis':402 'qualiti':224,238 'quot':554,563 'read':296,395,577 'readabl':594 'recent':636 'record':1002 'refer':410,647 'referenc':373 'regardless':390 'regul':374 'relat':41 'repli':700,714 'report':727 'request':217 'requir':340,973 'result':818 'revert':745 'review':399,667 'revis':270 'rewrit':196 'right':187 'risk':267,318,792,806 'rm':696 'rollback':771 'rule':634 'run':694,709 'said':486,780,826,884 'save':650,661,662 'scan':168 'secondari':687 'section':589,756 'see':449 'self':51,77,85,95,103,110,118,130,690,705,719,729 'self-improv':50,76,84,94,102,109,117,129,689,704,718,728 'sentenc':632,641 'separ':347 'session':11,42,74,80,157 'show':113,124 'signal':167,172,179,222,228,430,653,668,673 'singl':339,811,858 'skill':2,16,24,40,53,62,88,93,121,136,138,144,146,151,159,162,212,236,440,484,544,578,580,602,604,682,684,732,758,760,764,766 'skill-nam':87,120,543,579,603,683,731,759,765 'skill-optimizer-lawv':1 'skill-rel':39 'skill-skill-optimizer-lawvable' 'skill.md':576 'sourc':553,562 'source-lawvable' 'specif':92 'specifi':377,799,899 'split':345,368,984 'stabl':371,813,860 'standard':274,376,419,421,832,841,848,863,890,908 'state':938,958 'status':112,116,670,675,721 'step':132,164,218,426,450,452,522,532,538,570,648 'structur':262,500 'subject':291,850 'success':66,198 'suggest':32,269 'target':90 'term':266,292 'termin':401 'thorough':322 'three':366 'time':425,816,866 'tool':472,501,874 'top':617,638 'topic-agent-skills' 'topic-automation' 'topic-law' 'topic-legal-work' 'topic-workflows' 'touch':711 'trigger':75 'two':293 'type':180,855 'uncap':910 'unclear':970 'understand':300 'updat':516,575,601,753,763,943 'use':17,272,469,871 'user':31,47,176,195,202,208,487,513,526,779,825,941,950,965,1000 'user-fac':999 'vagu':289,994 'vari':852 'version':379,749 'want':925 'way':304,389 'without':204 'work':10,21,73 'workaround':211 'workflow':128 'would':299,741 'x':191 'y':194 'y/n':569 'yes':200,679,752,795,801,810,814,857","prices":[{"id":"128f7d85-6084-4518-a4c5-dddde7d26283","listingId":"c0b2e458-7c96-43fc-bfed-08bf0b2eec11","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"lawvable","category":"awesome-legal-skills","install_from":"skills.sh"},"createdAt":"2026-04-18T22:04:42.419Z"}],"sources":[{"listingId":"c0b2e458-7c96-43fc-bfed-08bf0b2eec11","source":"github","sourceId":"lawvable/awesome-legal-skills/skill-optimizer-lawvable","sourceUrl":"https://github.com/lawvable/awesome-legal-skills/tree/main/skills/skill-optimizer-lawvable","isPrimary":false,"firstSeenAt":"2026-04-18T22:04:42.419Z","lastSeenAt":"2026-05-02T18:53:56.780Z"}],"details":{"listingId":"c0b2e458-7c96-43fc-bfed-08bf0b2eec11","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"lawvable","slug":"skill-optimizer-lawvable","github":{"repo":"lawvable/awesome-legal-skills","stars":310,"topics":["agent-skills","automation","law","legal-work","workflows"],"license":"other","html_url":"https://github.com/lawvable/awesome-legal-skills","pushed_at":"2026-03-03T11:25:06Z","description":"A curated list of awesome Agent Skills for automating legal work","skill_md_sha":"65157fd99271e55314b710268f0a1fde02fd2872","skill_md_path":"skills/skill-optimizer-lawvable/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/lawvable/awesome-legal-skills/tree/main/skills/skill-optimizer-lawvable"},"layout":"multi","source":"github","category":"awesome-legal-skills","frontmatter":{"name":"skill-optimizer-lawvable","description":"Guide to analyze a current work session and propose improvements to skills. Use (1) automatically after working with a skill to capture learnings, (2) when the user suggests improvements, corrections, or additions during a skill-related session, or (3) when the user manually invokes `self-improve`."},"skills_sh_url":"https://skills.sh/lawvable/awesome-legal-skills/skill-optimizer-lawvable"},"updatedAt":"2026-05-02T18:53:56.780Z"}}