[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6G1ED8KGEXC19CBXMXDQPDC":3,"skill-same:01M6G1ED8KGEXC19CBXMXDQPDC":107},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":36,"linkOnly":37,"markdownContent":38,"markdownBody":39,"frontMatter":40,"refName":42,"commitSha":43,"skillDirectory":44,"forksCount":45,"repositoryHtmlUrl":46,"repositoryHomepageUrl":6,"ownerHtmlUrl":47,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":48,"install":49,"resources":62},"01M6G1ED8KGEXC19CBXMXDQPDC","token-saver",null,"Reduce OpenClaw AI costs with model-aware optimization. Features dynamic compaction presets based on your model's context window, intelligent file compression, and robust model detection with fallback. Supports Claude, GPT-4, Gemini, DeepSeek, and more.","kbarbel640-del\u002Fskills","kbarbel640-del","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F255315490?v=4","01M6CJPPQGCCESXMG3K7BBGQ5J","01M6CJPPX2PB2BCZERFVM2C2B3",1,"MIT","https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Frubenaquispe\u002Ftoken-saver\u002FSKILL.md",[],[18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35],"token saver","token","fallback","more","file compression","with","with model","compaction","deepseek","openclaw","context","model","detection","Optimization","compression","file","gpt","gemini","2026-11-10 10:12:30",false,"---\nname: token-saver\nversion: 3.0.0\ndescription: \"Reduce OpenClaw AI costs with model-aware optimization. Features dynamic compaction presets based on your model's context window, intelligent file compression, and robust model detection with fallback. Supports Claude, GPT-4, Gemini, DeepSeek, and more.\"\n---\n\n# Token Saver v3\n\n> **? Did you know?** Every API call sends your workspace files (SOUL.md, USER.md, MEMORY.md, AGENTS.md, etc.) along with your message. These files count toward your context window, slowing responses and costing real money on every message.\n\nToken Saver v3 is **model-aware** — it knows your model's context window and adapts recommendations accordingly. Using Gemini's 1M context? Presets scale up. On GPT-4o's 128K? Presets adjust down.\n\n## What's New in v3\n\n| Feature | v2 | v3 |\n|---------|----|----|\n| Compaction presets | Fixed (80K\u002F120K\u002F160K) | Dynamic (% of model's context) |\n| Model detection | Fragile, env-only | Robust fallback chain |\n| Context windows | Not tracked | Full registry (9 models) |\n| Model info | Hardcoded pricing | JSON registry, easy updates |\n| Already-optimized | Re-compressed | Smart bypass |\n\n## Commands\n\n| Command | What it does |\n|---|---|\n| `\u002Foptimize` | Full dashboard — files, models, context usage % |\n| `\u002Foptimize tokens` | Compress workspace files (auto-backup) |\n| `\u002Foptimize compaction` | Chat compaction control (model-aware) |\n| `\u002Foptimize compaction balanced` | Apply balanced preset (60% of context) |\n| `\u002Foptimize compaction 120` | Custom threshold (compact at 120K) |\n| `\u002Foptimize models` | Detailed model audit with registry |\n| `\u002Foptimize revert` | Restore backups, disable persistent mode |\n\n## Features\n\n### ? Model-Aware Dashboard\nShows current model, context window, and usage percentage:\n```\n? Model: Claude Opus 4.5 (200K context)\n   Detected: openclaw.json\n\n? Context Usage: [████████░░░░░░░░░░░░] 42% (84K\u002F200K)\n```\n\n### ? Workspace File Compression\nScans all `.md` files, shows token count and potential savings. Smart bypass skips already-optimized files.\n\n**File-aware compression:**\n- **SOUL.md** — Light compression, keeps personality language\n- **AGENTS.md** — Medium compression, dense instructions\n- **USER.md \u002F MEMORY.md** — Heavy compression, key:value format\n- **PROJECTS.md** — No compression (user structure preserved)\n\n### ? Dynamic Compaction Presets\nPresets adapt to your model's context window:\n\n| Preset | % of Context | Claude 200K | GPT-4o 128K | Gemini 1M |\n|--------|--------------|-------------|-------------|-----------|\n| Aggressive | 40% | 80K | 51K | 400K |\n| Balanced | 60% | 120K | 77K | 600K |\n| Conservative | 80% | 160K | 102K | 800K |\n| Off | 95% | 190K | 122K | 950K |\n\n### ? Model Registry\n24+ models with context windows, pricing, and aliases:\n- **Claude:** Opus 4.6 (1M), Opus 4.5, Sonnet 4.5, Sonnet 4, Haiku 4.5, Haiku 3.5 (200K)\n- **OpenAI:** GPT-5.2, GPT-5.1, GPT-5-mini, GPT-5-nano (256K), GPT-4.1, GPT-4o (128K), o1, o3, o4-mini\n- **Gemini:** 3 Pro (2M), 2.5 Pro, 2.0 Flash (1M)\n- **Others:** DeepSeek V3 (64K), Kimi K2.5 (128K), Llama 3.3 70B, Mistral Large\n\n### ? Robust Model Detection\nDetection priority:\n1. Runtime injection (`--model=...`)\n2. Environment variables (`SKILL_MODEL`, `OPENCLAW_MODEL`)\n3. Config file (`~\u002F.openclaw\u002Fopenclaw.json`)\n4. File inference (TOOLS.md, MEMORY.md mentions)\n5. Fallback: Claude Sonnet 4 (safe default)\n\n**Unknown model handling:**\n- Strict version matching — `opus-6.5` won't fuzzy-match to `opus-4.5`\n- Unknown models get safe defaults (200K context) + warning\n- Easy to add new models to `scripts\u002Fmodels.json`\n\n### ? Persistent Mode\nAdds writing guidance to AGENTS.md for continued token efficiency:\n\n| File | Writing Style |\n|------|---------------|\n| SOUL.md | Evocative, personality-shaping |\n| AGENTS.md | Dense instructions, symbols OK |\n| USER.md | Key:value facts |\n| MEMORY.md | Ultra-dense data |\n\n## Safety\n\n- **Auto-backup** — All modified files get `.backup` extension\n- **Integrity > Size** — Never sacrifices meaning for smaller tokens\n- **Smart bypass** — Skips already-optimized files\n- **Revert anytime** — `\u002Foptimize revert` restores everything\n- **No external calls** — All analysis runs locally\n\n## Installation\n\n```\nclawhub install token-saver --registry \"https:\u002F\u002Fwww.clawhub.ai\"\n```\n\n## Version History\n- **3.0.0** — Model registry, dynamic presets, robust detection, smart bypass\n- **2.0.1** — Chat compaction, file-aware compression, persistent mode\n- **1.0.0** — Initial release\n","# Token Saver v3\n\n> **? Did you know?** Every API call sends your workspace files (SOUL.md, USER.md, MEMORY.md, AGENTS.md, etc.) along with your message. These files count toward your context window, slowing responses and costing real money on every message.\n\nToken Saver v3 is **model-aware** — it knows your model's context window and adapts recommendations accordingly. Using Gemini's 1M context? Presets scale up. On GPT-4o's 128K? Presets adjust down.\n\n## What's New in v3\n\n| Feature | v2 | v3 |\n|---------|----|----|\n| Compaction presets | Fixed (80K\u002F120K\u002F160K) | Dynamic (% of model's context) |\n| Model detection | Fragile, env-only | Robust fallback chain |\n| Context windows | Not tracked | Full registry (9 models) |\n| Model info | Hardcoded pricing | JSON registry, easy updates |\n| Already-optimized | Re-compressed | Smart bypass |\n\n## Commands\n\n| Command | What it does |\n|---|---|\n| `\u002Foptimize` | Full dashboard — files, models, context usage % |\n| `\u002Foptimize tokens` | Compress workspace files (auto-backup) |\n| `\u002Foptimize compaction` | Chat compaction control (model-aware) |\n| `\u002Foptimize compaction balanced` | Apply balanced preset (60% of context) |\n| `\u002Foptimize compaction 120` | Custom threshold (compact at 120K) |\n| `\u002Foptimize models` | Detailed model audit with registry |\n| `\u002Foptimize revert` | Restore backups, disable persistent mode |\n\n## Features\n\n### ? Model-Aware Dashboard\nShows current model, context window, and usage percentage:\n```\n? Model: Claude Opus 4.5 (200K context)\n   Detected: openclaw.json\n\n? Context Usage: [████████░░░░░░░░░░░░] 42% (84K\u002F200K)\n```\n\n### ? Workspace File Compression\nScans all `.md` files, shows token count and potential savings. Smart bypass skips already-optimized files.\n\n**File-aware compression:**\n- **SOUL.md** — Light compression, keeps personality language\n- **AGENTS.md** — Medium compression, dense instructions\n- **USER.md \u002F MEMORY.md** — Heavy compression, key:value format\n- **PROJECTS.md** — No compression (user structure preserved)\n\n### ? Dynamic Compaction Presets\nPresets adapt to your model's context window:\n\n| Preset | % of Context | Claude 200K | GPT-4o 128K | Gemini 1M |\n|--------|--------------|-------------|-------------|-----------|\n| Aggressive | 40% | 80K | 51K | 400K |\n| Balanced | 60% | 120K | 77K | 600K |\n| Conservative | 80% | 160K | 102K | 800K |\n| Off | 95% | 190K | 122K | 950K |\n\n### ? Model Registry\n24+ models with context windows, pricing, and aliases:\n- **Claude:** Opus 4.6 (1M), Opus 4.5, Sonnet 4.5, Sonnet 4, Haiku 4.5, Haiku 3.5 (200K)\n- **OpenAI:** GPT-5.2, GPT-5.1, GPT-5-mini, GPT-5-nano (256K), GPT-4.1, GPT-4o (128K), o1, o3, o4-mini\n- **Gemini:** 3 Pro (2M), 2.5 Pro, 2.0 Flash (1M)\n- **Others:** DeepSeek V3 (64K), Kimi K2.5 (128K), Llama 3.3 70B, Mistral Large\n\n### ? Robust Model Detection\nDetection priority:\n1. Runtime injection (`--model=...`)\n2. Environment variables (`SKILL_MODEL`, `OPENCLAW_MODEL`)\n3. Config file (`~\u002F.openclaw\u002Fopenclaw.json`)\n4. File inference (TOOLS.md, MEMORY.md mentions)\n5. Fallback: Claude Sonnet 4 (safe default)\n\n**Unknown model handling:**\n- Strict version matching — `opus-6.5` won't fuzzy-match to `opus-4.5`\n- Unknown models get safe defaults (200K context) + warning\n- Easy to add new models to `scripts\u002Fmodels.json`\n\n### ? Persistent Mode\nAdds writing guidance to AGENTS.md for continued token efficiency:\n\n| File | Writing Style |\n|------|---------------|\n| SOUL.md | Evocative, personality-shaping |\n| AGENTS.md | Dense instructions, symbols OK |\n| USER.md | Key:value facts |\n| MEMORY.md | Ultra-dense data |\n\n## Safety\n\n- **Auto-backup** — All modified files get `.backup` extension\n- **Integrity > Size** — Never sacrifices meaning for smaller tokens\n- **Smart bypass** — Skips already-optimized files\n- **Revert anytime** — `\u002Foptimize revert` restores everything\n- **No external calls** — All analysis runs locally\n\n## Installation\n\n```\nclawhub install token-saver --registry \"https:\u002F\u002Fwww.clawhub.ai\"\n```\n\n## Version History\n- **3.0.0** — Model registry, dynamic presets, robust detection, smart bypass\n- **2.0.1** — Chat compaction, file-aware compression, persistent mode\n- **1.0.0** — Initial release\n",{"name":5,"version":41,"description":7},"3.0.0","main","e7e336ac628c3e3bd771aff1aaa469c93778f94a","skills\u002Frubenaquispe\u002Ftoken-saver",0,"https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills","https:\u002F\u002Fgithub.com\u002Fkbarbel640-del","2026-08-23 22:39:37",{"source":8,"skillName":5,"commands":50,"prompt":60,"zipAvailable":61},[51,54,57],{"tool":52,"command":53},"npx","npx skills add kbarbel640-del\u002Fskills --skill token-saver",{"tool":55,"command":56},"bunx","bunx skills add kbarbel640-del\u002Fskills --skill token-saver",{"tool":58,"command":59},"pnpm","pnpm dlx skills add kbarbel640-del\u002Fskills --skill token-saver","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Frubenaquispe\u002Ftoken-saver\u002FSKILL.md",true,[63,69,73,77,81,87,91,95,99,103],{"path":64,"resourceType":65,"resourceTypeDesc":66,"sizeBytes":67,"sourceHtmlUrl":68},"_meta.json","REFERENCE","参考文档",634,"https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Frubenaquispe\u002Ftoken-saver\u002F_meta.json",{"path":70,"resourceType":65,"resourceTypeDesc":66,"sizeBytes":71,"sourceHtmlUrl":72},"audit.json",6197,"https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Frubenaquispe\u002Ftoken-saver\u002Faudit.json",{"path":74,"resourceType":65,"resourceTypeDesc":66,"sizeBytes":75,"sourceHtmlUrl":76},"package.json",701,"https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Frubenaquispe\u002Ftoken-saver\u002Fpackage.json",{"path":78,"resourceType":65,"resourceTypeDesc":66,"sizeBytes":79,"sourceHtmlUrl":80},"README.md",3792,"https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Frubenaquispe\u002Ftoken-saver\u002FREADME.md",{"path":82,"resourceType":83,"resourceTypeDesc":84,"sizeBytes":85,"sourceHtmlUrl":86},"scripts\u002Fanalyzer.js","SCRIPT","脚本",17040,"https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Frubenaquispe\u002Ftoken-saver\u002Fscripts\u002Fanalyzer.js",{"path":88,"resourceType":83,"resourceTypeDesc":84,"sizeBytes":89,"sourceHtmlUrl":90},"scripts\u002Fcompressor.js",17156,"https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Frubenaquispe\u002Ftoken-saver\u002Fscripts\u002Fcompressor.js",{"path":92,"resourceType":83,"resourceTypeDesc":84,"sizeBytes":93,"sourceHtmlUrl":94},"scripts\u002Fmodels.json",7609,"https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Frubenaquispe\u002Ftoken-saver\u002Fscripts\u002Fmodels.json",{"path":96,"resourceType":83,"resourceTypeDesc":84,"sizeBytes":97,"sourceHtmlUrl":98},"scripts\u002Foptimizer.js",26607,"https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Frubenaquispe\u002Ftoken-saver\u002Fscripts\u002Foptimizer.js",{"path":100,"resourceType":65,"resourceTypeDesc":66,"sizeBytes":101,"sourceHtmlUrl":102},"v3-plan.md",5983,"https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Frubenaquispe\u002Ftoken-saver\u002Fv3-plan.md",{"path":104,"resourceType":65,"resourceTypeDesc":66,"sizeBytes":105,"sourceHtmlUrl":106},"v3-test-results.md",3799,"https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Frubenaquispe\u002Ftoken-saver\u002Fv3-test-results.md",[108,138,166,187,193,213],{"skillNo":109,"name":110,"nameZh":6,"description":111,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":112,"categories":113,"keywords":119,"publishedAt":36},"01M6G8J55967D6281ACMKAKTEQ","crowd-prompting","A marketplace where AI agents improve prompts, system instructions, tool descriptions, and other text-based content with domain expertise from real-world operations — and earn tokens for valuable contributions.","https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fzzadrian\u002Fcrowd-prompting\u002FSKILL.md",[114],{"categoryNo":115,"categoryName":116,"taxonomyType":117,"taxonomyTypeDesc":118},"01JBSEEDCAT000000000000036","内容创作者","PROFESSION","适用职业",[120,121,122,23,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137],"crowd prompting","prompting","api","tokens","domain","world","prompts","operations","instructions","text","tool","system","improve","productivity","marketplace","https","content","ai agents",{"skillNo":139,"name":140,"nameZh":6,"description":141,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":142,"categories":143,"keywords":149,"publishedAt":36},"01M6G8J0PKPAYVET6BWE8W4C26","dilemm-ai","Compete in DilemmAI, the prisoner's dilemma AI arena at dilemm.ai. Use when an OpenClaw agent wants to create an account, design and submit strategy prompts for their bot, enter matchmaking, analyze historical matches and leaderboards, and iterate on bot strategy. Requires headless browser automation (no API available yet).","https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fzyori\u002Fdilemm-ai\u002FSKILL.md",[144],{"categoryNo":145,"categoryName":146,"taxonomyType":147,"taxonomyTypeDesc":148},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[150,122,151,152,153,154,155,156,157,158,27,159,160,161,162,163,164,126,165],"dilemm ai","arena","analyze","browser","openclaw agent","create","automation","browser automation","bot","design","strategy","headless","compete","submit","matchmaking","use",{"skillNo":167,"name":168,"nameZh":6,"description":169,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":170,"categories":171,"keywords":175,"publishedAt":36},"01M6G8HWAF3K9SW7NP3Z83CPJ6","arxiv-paper-reviews","Interact with arXiv Crawler API to fetch papers, read reviews, and submit comments. Use when working with arXiv papers, fetching paper lists by date\u002Fcategory\u002Finterest, viewing paper details with comments, or submitting paper reviews via API at http:\u002F\u002F150.158.152.82:8000.","https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fzxrys\u002Fweak-accept\u002FSKILL.md",[172],{"categoryNo":173,"categoryName":174,"taxonomyType":147,"taxonomyTypeDesc":148},"01JBSEEDCAT000000000000021","资料检索与研究",[176,177,178,179,180,181,182,23,183,163,165,184,122,185,186],"arxiv paper reviews","reviews","arxiv","paper","crawler","papers","comments","fetch","read","http","via",{"skillNo":188,"name":168,"nameZh":6,"description":169,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":189,"categories":190,"keywords":192,"publishedAt":36},"01M6G8HQJ7N7WMY6YV7D7SSJ16","https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fzxrys\u002Farxiv-paper-reviews\u002FSKILL.md",[191],{"categoryNo":173,"categoryName":174,"taxonomyType":147,"taxonomyTypeDesc":148},[179,178,177,176,186,181,182,23,180,163,165,183,184,122,185],{"skillNo":194,"name":195,"nameZh":6,"description":196,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":197,"categories":198,"keywords":199,"publishedAt":36},"01M6G8HM95Y64JNVFC2HN4E2BG","git-summary","Get a quick summary of the current Git repository including status, recent commits, branches, and contributors.","https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fzweack\u002Fgit-summary\u002FSKILL.md",[],[200,201,202,203,204,205,206,207,208,209,210,211,212],"git","summary","git summary","commits","repository","win32","status","linux","contributors","true","quick","get","branches",{"skillNo":214,"name":215,"nameZh":6,"description":216,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":217,"categories":218,"keywords":223,"publishedAt":36},"01M6G8HGEGF05CNEJCGVYFSC4R","agent-constitution","Interact with AgentConstitution governance contracts on Base Sepolia. Check compliance, read rules, log actions, query governance state.","https:\u002F\u002Fgithub.com\u002Fkbarbel640-del\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fztsalexey\u002Fagent-constitution\u002FSKILL.md",[219,220],{"categoryNo":145,"categoryName":146,"taxonomyType":147,"taxonomyTypeDesc":148},{"categoryNo":221,"categoryName":222,"taxonomyType":117,"taxonomyTypeDesc":118},"01JBSEEDCAT000000000000037","律师与法务",[224,225,226,227,23,228,229,42,230,184,231,232,233,234,235,135,236,237,238],"agent constitution","constitution","contracts","tree","query","base","log","check","Rules","projects","compliance","actions","state","governance","github"]