[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"extension-skill-juliusbrussee-caveman-compress-en":3,"guides-for-juliusbrussee-caveman-compress":500,"similar-k176a757cfd35jb9npje9qqzed86nvpn-en":501},{"_creationTime":4,"_id":5,"children":6,"community":7,"display":9,"evaluation":15,"identity":243,"isFallback":225,"parentExtension":248,"providers":302,"relations":306,"repo":307,"tags":497,"workflow":498},1778691765524.3738,"k176a757cfd35jb9npje9qqzed86nvpn",[],{"reviewCount":8},0,{"description":10,"installMethods":11,"name":13,"sourceUrl":14},"Compress natural language memory files (CLAUDE.md, todos, preferences) into caveman format to save input tokens. Preserves all technical substance, code, URLs, and structure. Compressed version overwrites the original file. Human-readable backup saved as FILE.original.md. Trigger: /caveman-compress FILEPATH or \"compress memory file\"\n",{"claudeCode":12},"juliusbrussee/caveman","caveman-compress","https://github.com/juliusbrussee/caveman",{"_creationTime":16,"_id":17,"extensionId":5,"locale":18,"result":19,"trustSignals":223,"workflow":241},1778691883875.7493,"kn7ce6nzqeewkc3j831hecgeqn86n7tj","en",{"checks":20,"evaluatedAt":192,"extensionSummary":193,"features":194,"nonGoals":200,"promptVersionExtension":205,"promptVersionScoring":206,"purpose":207,"rationale":208,"score":209,"summary":210,"tags":211,"targetMarket":217,"tier":218,"useCases":219},[21,26,29,32,36,39,44,48,51,54,58,62,65,69,72,75,78,81,84,87,91,95,99,103,107,110,113,116,120,123,126,129,132,135,138,142,146,150,153,157,160,163,166,169,173,176,179,182,185,189],{"category":22,"check":23,"severity":24,"summary":25},"Practical Utility","Problem relevance","pass","The description clearly identifies the problem of saving input tokens by compressing memory files and names the specific artifacts (.md, todos, preferences) and the user intent (saving tokens, reducing cost).",{"category":22,"check":27,"severity":24,"summary":28},"Unique selling proposition","The skill offers significant value over default LLM behavior by providing a specialized compression algorithm for natural language files, preserving technical content while reducing token count, which goes beyond simple summarization or rephrasing.",{"category":22,"check":30,"severity":24,"summary":31},"Production readiness","The skill is production-ready, addressing the complete lifecycle of file compression: it detects file types, compresses, validates, handles errors with retries, backs up originals, and overwrites the compressed file, all covered by tests.",{"category":33,"check":34,"severity":24,"summary":35},"Scope","Single responsibility principle","The skill focuses solely on compressing natural language files into a more token-efficient format, with clear boundaries around what it will and will not compress.",{"category":33,"check":37,"severity":24,"summary":38},"Description quality","The displayed description accurately reflects the skill's functionality, including its trigger, purpose, backup mechanism, and file preservation rules.",{"category":40,"check":41,"severity":42,"summary":43},"Invocation","Scoped tools","not_applicable","This skill is a single command-line tool and does not expose multiple distinct tools that could be evaluated for scoping.",{"category":45,"check":46,"severity":42,"summary":47},"Documentation","Configuration & parameter reference","The skill does not have configurable parameters or environment variables beyond the Anthropic API key, which is handled by the environment or CLI configuration.",{"category":33,"check":49,"severity":24,"summary":50},"Tool naming","The skill's primary command is `/caveman-compress` which is descriptive of its function.",{"category":33,"check":52,"severity":24,"summary":53},"Minimal I/O surface","The input is a single file path, and the output is the modified file and a backup, adhering to the minimal required I/O for its task.",{"category":55,"check":56,"severity":24,"summary":57},"License","License usability","The extension is distributed under the MIT license, as indicated by the bundled LICENSE file, which is a permissive open-source license.",{"category":59,"check":60,"severity":24,"summary":61},"Maintenance","Commit recency","The latest commit was on May 12, 2026, which is within the last 3 months.",{"category":59,"check":63,"severity":24,"summary":64},"Dependency Management","The skill uses Python and relies on the `anthropic` library and standard Python libraries. Python's packaging and environment management are sufficient for this skill.",{"category":66,"check":67,"severity":24,"summary":68},"Security","Secret Management","The skill handles API keys via environment variables or the CLI's auth, and its SECURITY.md explicitly states it does not echo secrets and refuses to compress sensitive file types.",{"category":66,"check":70,"severity":24,"summary":71},"Injection","The skill is designed to treat file content as data and explicitly avoids executing it as code. It handles sensitive paths by refusing to process them.",{"category":66,"check":73,"severity":24,"summary":74},"Transitive Supply-Chain Grenades","The skill processes local files and calls the Anthropic API directly or via the CLI, with no runtime fetching of external code or instructions.",{"category":66,"check":76,"severity":24,"summary":77},"Sandbox Isolation","The skill operates only on the specified input file and writes to the same location, respecting sandbox boundaries.",{"category":66,"check":79,"severity":24,"summary":80},"Sandbox escape primitives","The Python scripts use standard library calls and `subprocess.run` without shell=True, avoiding common sandbox escape primitives.",{"category":66,"check":82,"severity":24,"summary":83},"Data Exfiltration","The skill is designed to avoid exfiltrating confidential data by refusing to process sensitive file types and only interacting with the Anthropic API for compression.",{"category":66,"check":85,"severity":24,"summary":86},"Hidden Text Tricks","The bundled scripts and markdown files do not contain hidden text tricks or obfuscated content.",{"category":88,"check":89,"severity":24,"summary":90},"Hooks","Opaque code execution","The Python scripts are provided as clear, readable source code without obfuscation, base64 encoding, or runtime downloads.",{"category":92,"check":93,"severity":24,"summary":94},"Portability","Structural Assumption","The script operates on a user-provided file path and does not make assumptions about the project's directory structure.",{"category":96,"check":97,"severity":24,"summary":98},"Trust","Issues Attention","With 66 issues opened and 70 closed in the last 90 days, the closure rate is approximately 51%, indicating good maintainer engagement.",{"category":100,"check":101,"severity":24,"summary":102},"Versioning","Release Management","The `scripts/__init__.py` file declares a `__version__ = \"1.0.0\"`, indicating a detectable version for the script package.",{"category":104,"check":105,"severity":24,"summary":106},"Code Execution","Validation","The validation script checks for exact preservation of headings, code blocks, URLs, paths, and inline code, ensuring integrity.",{"category":66,"check":108,"severity":24,"summary":109},"Unguarded Destructive Operations","The skill overwrites the original file but creates a backup first, and it does not perform inherently destructive operations beyond this controlled modification.",{"category":104,"check":111,"severity":24,"summary":112},"Error Handling","The `compress.py` script includes retries, checks for Claude failures, empty responses, identical output, and restores the original file on repeated validation failures, reporting errors clearly.",{"category":104,"check":114,"severity":24,"summary":115},"Logging","The script provides console output for processing steps, validation attempts, and success/failure messages, serving as a basic audit trail.",{"category":117,"check":118,"severity":24,"summary":119},"Compliance","GDPR","The skill refuses to compress files with sensitive names or paths, preventing the accidental submission of PII or credentials to the Anthropic API.",{"category":117,"check":121,"severity":24,"summary":122},"Target market","The skill operates on local files and interacts with the Anthropic API, with no regional or jurisdictional limitations detected; the target market is global.",{"category":92,"check":124,"severity":24,"summary":125},"Runtime stability","The script specifies Python 3.10+ and uses standard libraries, ensuring broad POSIX compatibility and graceful fallback to the CLI if the SDK is unavailable.",{"category":45,"check":127,"severity":24,"summary":128},"README","The README provides a clear overview, installation instructions, benchmarks, and usage examples, effectively stating the extension's purpose.",{"category":33,"check":130,"severity":42,"summary":131},"Tool surface size","This is a single-purpose skill invoked via a command-line interface, not a set of multiple tools.",{"category":40,"check":133,"severity":42,"summary":134},"Overlapping near-synonym tools","The skill exposes a single primary command, so there are no overlapping tools.",{"category":45,"check":136,"severity":24,"summary":137},"Phantom features","All advertised features, such as compression of specific file types and preservation rules, are implemented and verifiable in the code and documentation.",{"category":139,"check":140,"severity":24,"summary":141},"Install","Installation instruction","The README provides clear installation instructions for the `caveman` plugin and usage examples for the `/caveman-compress` command.",{"category":143,"check":144,"severity":24,"summary":145},"Errors","Actionable error messages","Errors are reported with context (e.g., file not found, too large, sensitive path, validation failure) and often include remediation steps or clear reasons for failure.",{"category":147,"check":148,"severity":24,"summary":149},"Execution","Pinned dependencies","The script specifies `python3` and `#!/usr/bin/env python3`, and while specific library versions aren't pinned in a lockfile, the use of standard Python libraries makes it portable.",{"category":33,"check":151,"severity":42,"summary":152},"Dry-run preview","The skill's core operation is file modification. A dry-run isn't applicable as the primary action is the compression and overwrite.",{"category":154,"check":155,"severity":42,"summary":156},"Protocol","Idempotent retry & timeouts","The operation is on local files, not remote calls, and while retries are implemented for LLM calls, timeouts are not explicitly configured for local operations.",{"category":117,"check":158,"severity":24,"summary":159},"Telemetry opt-in","The skill does not emit any telemetry, fulfilling the opt-in requirement by default.",{"category":40,"check":161,"severity":24,"summary":162},"Precise Purpose","The skill clearly states it compresses natural language files (CLAUDE.md, todos, preferences) into caveman format to save input tokens, with specific triggers and explicit boundaries.",{"category":40,"check":164,"severity":24,"summary":165},"Concise Frontmatter","The frontmatter in SKILL.md is concise, clearly stating the name, description, trigger, and purpose within a reasonable character limit.",{"category":45,"check":167,"severity":24,"summary":168},"Concise Body","The SKILL.md body is concise, outlining the process and rules, and delegates detailed explanations to separate files or inline documentation.",{"category":170,"check":171,"severity":24,"summary":172},"Context","Progressive Disclosure","The SKILL.md is concise, and while it doesn't explicitly link to separate `reference/` files, it provides sufficient detail inline without embedding large blobs.",{"category":170,"check":174,"severity":42,"summary":175},"Forked exploration","This skill is a direct file manipulation tool and does not involve deep exploration or code review requiring a forked context.",{"category":22,"check":177,"severity":24,"summary":178},"Usage examples","The README provides clear usage examples with file paths and the SKILL.md details the command structure, demonstrating how to invoke the skill.",{"category":22,"check":180,"severity":24,"summary":181},"Edge cases","The skill documents and handles edge cases like empty files, large files, sensitive filenames, identical output, and validation failures, with recovery steps documented.",{"category":104,"check":183,"severity":24,"summary":184},"Tool Fallback","The script uses the `anthropic` SDK if the API key is set, falling back to the `claude` CLI for authentication, acting as a functional fallback.",{"category":186,"check":187,"severity":24,"summary":188},"Safety","Halt on unexpected state","The script halts compression on unexpected states like empty files, large files, sensitive paths, or if Claude returns unusable output, and restores the original file on persistent validation failures.",{"category":92,"check":190,"severity":24,"summary":191},"Cross-skill coupling","The skill is self-contained and does not implicitly rely on other skills being loaded in the same session.",1778691883697,"This skill compresses natural language files like .md, .txt, and .tex into a more token-efficient 'caveman' format. It preserves code blocks, URLs, and structure, overwrites the original file, and saves a human-readable backup. It uses the Anthropic SDK or CLI and includes robust validation and error handling.",[195,196,197,198,199],"Compresses natural language files to save tokens","Preserves code blocks, URLs, and file paths exactly","Overwrites original file with compressed version","Saves a human-readable backup of the original file","Handles validation, retries, and error recovery",[201,202,203,204],"Compressing code files (.py, .js, .json, etc.)","Compressing configuration files (.yaml, .env, .toml)","Modifying or deleting original files without creating a backup","Compressing sensitive files like credentials or private keys","3.0.0","4.4.0","Save Claude input tokens and reduce session costs by compressing natural language memory files into a more concise format.","The extension demonstrates excellent code quality, robust error handling, and comprehensive documentation. It has a very low score ceiling due to a single 'pass' on the 'Dependency Management' check, which is not applicable to its use of standard libraries.",99,"A high-quality, production-ready skill for compressing natural language files to save Claude tokens.",[212,213,214,215,216],"compression","token-saving","markdown","text-processing","utility","global","verified",[220,221,222],"Reduce Claude session costs by compressing large CLAUDE.md or preference files","Make project notes and todo lists more token-efficient for frequent loading","Save input tokens without losing technical substance or code",{"codeQuality":224,"collectedAt":226,"documentation":227,"maintenance":230,"popularity":236,"security":237,"testCoverage":240},{"hasLockfile":225},false,1778691864302,{"descriptionLength":228,"readmeSize":229},335,10896,{"closedIssues90d":231,"forks":232,"hasChangelog":225,"openIssues90d":233,"pushedAt":234,"stars":235},70,3292,66,1778615088000,59509,{"npmDownloads":8},{"hasNpmPackage":238,"license":239,"smitheryVerified":225},true,"MIT",{"hasCi":238,"hasTests":238},{"updatedAt":242},1778691883875,{"basePath":244,"githubOwner":245,"githubRepo":246,"locale":18,"slug":13,"type":247},"skills/caveman-compress","juliusbrussee","caveman","skill",{"_creationTime":249,"_id":250,"community":251,"display":252,"identity":256,"parentExtension":259,"providers":290,"relations":297,"tags":298,"workflow":299},1778691765524.373,"k170g496f5ze5hv8xa2e7b91ex86mz9m",{"reviewCount":8},{"description":253,"installMethods":254,"name":255,"sourceUrl":14},"Ultra-compressed communication mode. Cuts ~75% of tokens while keeping full technical accuracy by speaking like a caveman.",{"claudeCode":246},"Caveman",{"basePath":257,"githubOwner":245,"githubRepo":246,"locale":18,"slug":246,"type":258},"","plugin",{"_creationTime":260,"_id":261,"community":262,"display":263,"identity":266,"providers":268,"relations":283,"tags":285,"workflow":286},1778691765524.3728,"k17dcfwpwpzjc43zt2qawvft3h86mmse",{"reviewCount":8},{"description":264,"installMethods":265,"name":246,"sourceUrl":14},"Ultra-compressed communication mode for Claude Code. Cuts ~75% of tokens while keeping full technical accuracy.",{"claudeCode":12},{"basePath":257,"githubOwner":245,"githubRepo":246,"locale":18,"slug":246,"type":267},"marketplace",{"evaluate":269,"extract":277},{"promptVersionExtension":270,"promptVersionScoring":206,"score":271,"tags":272,"targetMarket":217,"tier":218},"3.1.0",92,[212,273,274,275,276],"llm","productivity","token-reduction","developer-tools",{"commitSha":278,"marketplace":279,"plugin":281},"HEAD",{"name":246,"pluginCount":280},1,{"mcpCount":8,"provider":282,"skillCount":8},"classify",{"repoId":284},"kd7890f0sfy0ymf1pakpx7vqch86nmew",[212,276,273,274,275],{"evaluatedAt":287,"extractAt":288,"updatedAt":289},1778691801317,1778691765524,1778692173865,{"evaluate":291,"extract":294},{"promptVersionExtension":205,"promptVersionScoring":206,"score":209,"tags":292,"targetMarket":217,"tier":218},[212,273,293,276,274],"text",{"commitSha":278,"license":239,"plugin":295},{"mcpCount":8,"provider":282,"skillCount":296},6,{"parentExtensionId":261,"repoId":284},[212,276,273,274,293],{"evaluatedAt":300,"extractAt":288,"updatedAt":301},1778691819811,1778692174193,{"evaluate":303,"extract":305},{"promptVersionExtension":205,"promptVersionScoring":206,"score":209,"tags":304,"targetMarket":217,"tier":218},[212,213,214,215,216],{"commitSha":278},{"parentExtensionId":250,"repoId":284},{"_creationTime":308,"_id":284,"identity":309,"providers":310,"workflow":492},1778691760140.7458,{"githubOwner":245,"githubRepo":246,"sourceUrl":14},{"classify":311,"discover":476,"extract":479,"github":480,"npm":491},{"commitSha":278,"extensions":312},[313,323,334,378,387,395,417,426,435,444,453,465],{"basePath":314,"displayName":315,"installMethods":316,"rationale":317,"selectedPaths":318,"source":322,"sourceLanguage":18,"type":267},".agents/plugins","caveman-repo",{"claudeCode":12},"marketplace.json at .agents/plugins/marketplace.json",[319],{"path":320,"priority":321},"marketplace.json","mandatory","rule",{"basePath":257,"description":264,"displayName":246,"installMethods":324,"rationale":325,"selectedPaths":326,"source":322,"sourceLanguage":18,"type":267},{"claudeCode":12},"marketplace.json at .claude-plugin/marketplace.json",[327,329,331],{"path":328,"priority":321},".claude-plugin/marketplace.json",{"path":330,"priority":321},"README.md",{"path":332,"priority":333},"LICENSE","high",{"basePath":257,"description":253,"displayName":246,"installMethods":335,"rationale":336,"selectedPaths":337,"source":322,"sourceLanguage":18,"type":258},{"claudeCode":246},"plugin manifest at .claude-plugin/plugin.json (coalesced with duplicate plugin at plugins/caveman)",[338,340,341,342,345,347,349,351,353,355,357,359,361,363,366,368,370,372,374,376],{"path":339,"priority":321},".claude-plugin/plugin.json",{"path":330,"priority":321},{"path":332,"priority":333},{"path":343,"priority":344},"skills/cavecrew/SKILL.md","medium",{"path":346,"priority":344},"skills/caveman/SKILL.md",{"path":348,"priority":344},"skills/caveman-commit/SKILL.md",{"path":350,"priority":344},"skills/caveman-compress/SKILL.md",{"path":352,"priority":344},"skills/caveman-help/SKILL.md",{"path":354,"priority":344},"skills/caveman-review/SKILL.md",{"path":356,"priority":344},"skills/caveman-stats/SKILL.md",{"path":358,"priority":333},"agents/cavecrew-builder.md",{"path":360,"priority":333},"agents/cavecrew-investigator.md",{"path":362,"priority":333},"agents/cavecrew-reviewer.md",{"path":364,"priority":365},"bin/install.js","low",{"path":367,"priority":365},"bin/lib/openclaw.js",{"path":369,"priority":365},"bin/lib/settings.js",{"path":371,"priority":333},"commands/caveman-commit.toml",{"path":373,"priority":333},"commands/caveman-init.toml",{"path":375,"priority":333},"commands/caveman-review.toml",{"path":377,"priority":333},"commands/caveman.toml",{"basePath":379,"description":380,"displayName":381,"installMethods":382,"rationale":383,"selectedPaths":384,"source":322,"sourceLanguage":18,"type":247},".junie/skills/cavecrew","Decision guide for delegating to caveman-style subagents. Tells the main thread WHEN to spawn `cavecrew-investigator` (locate code), `cavecrew-builder` (1-2 file edit), or `cavecrew-reviewer` (diff review) instead of doing the work inline or using vanilla `Explore`. Subagent output is caveman-compressed so the tool-result injected back into main context is ~60% smaller — main context lasts longer across long sessions. Trigger: \"delegate to subagent\", \"use cavecrew\", \"spawn investigator/builder/reviewer\", \"save context\", \"compressed agent output\".\n","cavecrew",{"claudeCode":12},"SKILL.md frontmatter at .junie/skills/cavecrew/SKILL.md (coalesced with duplicate skill at .agents/skills/cavecrew, skills/cavecrew, .roo/skills/cavecrew, .kiro/skills/cavecrew, plugins/caveman/skills/cavecrew)",[385],{"path":386,"priority":321},"SKILL.md",{"basePath":388,"description":389,"displayName":246,"installMethods":390,"rationale":391,"selectedPaths":392,"source":322,"sourceLanguage":18,"type":247},"skills/caveman","Ultra-compressed communication mode. Cuts token usage ~75% by speaking like caveman while keeping full technical accuracy. Supports intensity levels: lite, full (default), ultra, wenyan-lite, wenyan-full, wenyan-ultra. Use when user says \"caveman mode\", \"talk like caveman\", \"use caveman\", \"less tokens\", \"be brief\", or invokes /caveman. Also auto-triggers when token efficiency is requested.\n",{"claudeCode":12},"SKILL.md frontmatter at skills/caveman/SKILL.md (coalesced with duplicate skill at plugins/caveman/skills/caveman)",[393,394],{"path":386,"priority":321},{"path":330,"priority":333},{"basePath":244,"description":10,"displayName":13,"installMethods":396,"rationale":397,"selectedPaths":398,"source":322,"sourceLanguage":18,"type":247},{"claudeCode":12},"SKILL.md frontmatter at skills/caveman-compress/SKILL.md (coalesced with duplicate skill at plugins/caveman/skills/caveman-compress)",[399,400,401,403,405,407,409,411,413,415],{"path":386,"priority":321},{"path":330,"priority":333},{"path":402,"priority":344},"SECURITY.md",{"path":404,"priority":365},"scripts/__init__.py",{"path":406,"priority":365},"scripts/__main__.py",{"path":408,"priority":365},"scripts/benchmark.py",{"path":410,"priority":365},"scripts/cli.py",{"path":412,"priority":365},"scripts/compress.py",{"path":414,"priority":365},"scripts/detect.py",{"path":416,"priority":365},"scripts/validate.py",{"basePath":418,"description":419,"displayName":420,"installMethods":421,"rationale":422,"selectedPaths":423,"source":322,"sourceLanguage":18,"type":247},"skills/caveman-stats","Show real token usage and estimated savings for the current session. Reads directly from the Claude Code session log — no AI estimation. Triggers on /caveman-stats. Output is injected by the mode-tracker hook; the model itself does not compute the numbers.\n","caveman-stats",{"claudeCode":12},"SKILL.md frontmatter at skills/caveman-stats/SKILL.md (coalesced with duplicate skill at plugins/caveman/skills/caveman-stats)",[424,425],{"path":386,"priority":321},{"path":330,"priority":333},{"basePath":427,"description":428,"displayName":429,"installMethods":430,"rationale":431,"selectedPaths":432,"source":322,"sourceLanguage":18,"type":247},"skills/caveman-commit","Ultra-compressed commit message generator. Cuts noise from commit messages while preserving intent and reasoning. Conventional Commits format. Subject ≤50 chars, body only when \"why\" isn't obvious. Use when user says \"write a commit\", \"commit message\", \"generate commit\", \"/commit\", or invokes /caveman-commit. Auto-triggers when staging changes.\n","caveman-commit",{"claudeCode":12},"SKILL.md frontmatter at skills/caveman-commit/SKILL.md",[433,434],{"path":386,"priority":321},{"path":330,"priority":333},{"basePath":436,"description":437,"displayName":438,"installMethods":439,"rationale":440,"selectedPaths":441,"source":322,"sourceLanguage":18,"type":247},"skills/caveman-help","Quick-reference card for all caveman modes, skills, and commands. One-shot display, not a persistent mode. Trigger: /caveman-help, \"caveman help\", \"what caveman commands\", \"how do I use caveman\".\n","caveman-help",{"claudeCode":12},"SKILL.md frontmatter at skills/caveman-help/SKILL.md",[442,443],{"path":386,"priority":321},{"path":330,"priority":333},{"basePath":445,"description":446,"displayName":447,"installMethods":448,"rationale":449,"selectedPaths":450,"source":322,"sourceLanguage":18,"type":247},"skills/caveman-review","Ultra-compressed code review comments. Cuts noise from PR feedback while preserving the actionable signal. Each comment is one line: location, problem, fix. Use when user says \"review this PR\", \"code review\", \"review the diff\", \"/review\", or invokes /caveman-review. Auto-triggers when reviewing pull requests.\n","caveman-review",{"claudeCode":12},"SKILL.md frontmatter at skills/caveman-review/SKILL.md",[451,452],{"path":386,"priority":321},{"path":330,"priority":333},{"basePath":257,"description":454,"displayName":455,"installMethods":456,"license":239,"rationale":457,"selectedPaths":458,"source":322,"sourceLanguage":18,"type":464},"Caveman installer — detects your AI coding agents and installs caveman for each one.","caveman-installer",{"npm":455},"cli ecosystem detected at /",[459,461,462,463],{"path":460,"priority":321},"package.json",{"path":330,"priority":321},{"path":332,"priority":333},{"path":364,"priority":344},"cli",{"basePath":466,"description":467,"displayName":468,"installMethods":469,"license":239,"rationale":470,"selectedPaths":471,"source":322,"sourceLanguage":18,"type":464},"src/mcp-servers/caveman-shrink","MCP proxy that compresses prose fields (tool descriptions, etc.) using caveman rules. Same accuracy, fewer context tokens.","caveman-shrink",{"npm":468},"cli ecosystem detected at src/mcp-servers/caveman-shrink",[472,473,474],{"path":460,"priority":321},{"path":330,"priority":321},{"path":475,"priority":344},"index.js",{"sources":477},[478],"manual",{"npmPackage":455},{"closedIssues90d":231,"description":481,"forks":232,"homepage":482,"license":239,"openIssues90d":233,"pushedAt":234,"readmeSize":229,"stars":235,"topics":483},"🪨 why use many token when few token do trick — Claude Code skill that cuts 65% of tokens by talking like caveman","https://getcaveman.dev/",[484,485,246,486,487,273,488,489,247,490],"ai","anthropic","claude","claude-code","meme","prompt-engineering","tokens",{"downloads":8},{"classifiedAt":493,"discoverAt":494,"extractAt":495,"githubAt":495,"npmAt":496,"updatedAt":493},1778691765329,1778691760140,1778691762475,1778691764072,[212,214,215,213,216],{"evaluatedAt":242,"extractAt":288,"updatedAt":499},1778692174897,[],[502,530,561,590,618,647],{"_creationTime":503,"_id":504,"community":505,"display":506,"identity":512,"providers":516,"relations":523,"tags":526,"workflow":527},1778692846401.1528,"k17ah0hhhwvmsj6h6vxmfmd9nn86mhff",{"reviewCount":8},{"description":507,"installMethods":508,"name":510,"sourceUrl":511},"Ultra-compressed communication mode. Cuts token usage ~75% by dropping filler, articles, and pleasantries while keeping full technical accuracy. Use when user says \"caveman mode\", \"talk like caveman\", \"use caveman\", \"less tokens\", \"be brief\", or invokes /caveman.\n",{"claudeCode":509},"mattpocock/skills","Caveman Mode","https://github.com/mattpocock/skills",{"basePath":513,"githubOwner":514,"githubRepo":515,"locale":18,"slug":246,"type":247},"skills/productivity/caveman","mattpocock","skills",{"evaluate":517,"extract":522},{"promptVersionExtension":205,"promptVersionScoring":206,"score":518,"tags":519,"targetMarket":217,"tier":218},98,[520,212,521,273,293,216],"communication","efficiency",{"commitSha":278,"license":239},{"parentExtensionId":524,"repoId":525},"k171mc92f05dc4v6jzgf6bng7d86mr9t","kd70nm2vazehgqc4j28ve6cwv986mv1s",[520,212,521,273,293,216],{"evaluatedAt":528,"extractAt":529,"updatedAt":528},1778693459372,1778692846401,{"_creationTime":531,"_id":532,"community":533,"display":534,"identity":540,"providers":545,"relations":554,"tags":557,"workflow":558},1778694269038.6682,"k1752cypc448mke749yjbkc65186mg6f",{"reviewCount":8},{"description":535,"installMethods":536,"name":538,"sourceUrl":539},"This skill should be used when the user asks to \"compress context\", \"summarize conversation history\", \"implement compaction\", \"reduce token usage\", or mentions context compression, structured summarization, tokens-per-task optimization, or long-running agent sessions exceeding context limits.",{"claudeCode":537},"muratcankoylan/Agent-Skills-for-Context-Engineering","Context Compression","https://github.com/muratcankoylan/Agent-Skills-for-Context-Engineering",{"basePath":541,"githubOwner":542,"githubRepo":543,"locale":18,"slug":544,"type":247},"skills/context-compression","muratcankoylan","Agent-Skills-for-Context-Engineering","context-compression",{"evaluate":546,"extract":553},{"promptVersionExtension":205,"promptVersionScoring":206,"score":547,"tags":548,"targetMarket":217,"tier":218},100,[549,273,550,551,212,552],"context-engineering","agent","summarization","evaluation",{"commitSha":278,"license":239},{"parentExtensionId":555,"repoId":556},"k1754dy3wbsv2a5gr1a983zzs586njca","kd7f12maf5nxmx5xttjx7scfnx86m1tv",[550,212,549,552,273,551],{"evaluatedAt":559,"extractAt":560,"updatedAt":559},1778694410149,1778694269038,{"_creationTime":562,"_id":563,"community":564,"display":565,"identity":571,"providers":575,"relations":583,"tags":586,"workflow":587},1778691497595.9666,"k170aa7za6n5wvxtwx7xj8dc9d86mx9e",{"reviewCount":8},{"description":566,"installMethods":567,"name":569,"sourceUrl":570},"Compresses images to WebP (default) or PNG with automatic tool selection. Use when user asks to \"compress image\", \"optimize image\", \"convert to webp\", or reduce image file size.",{"claudeCode":568},"jimliu/baoyu-skills","baoyu-compress-image","https://github.com/jimliu/baoyu-skills",{"basePath":572,"githubOwner":573,"githubRepo":574,"locale":18,"slug":569,"type":247},"skills/baoyu-compress-image","jimliu","baoyu-skills",{"evaluate":576,"extract":582},{"promptVersionExtension":205,"promptVersionScoring":206,"score":547,"tags":577,"targetMarket":217,"tier":218},[578,212,579,580,581],"image","webp","png","optimization",{"commitSha":278},{"parentExtensionId":584,"repoId":585},"k172q4hxm2acmg5ree5aeemx1x86m6hx","kd718jn448p61w289hzf15tecd86n3a4",[212,578,581,580,579],{"evaluatedAt":588,"extractAt":589,"updatedAt":588},1778691619519,1778691497596,{"_creationTime":591,"_id":592,"community":593,"display":594,"identity":600,"providers":603,"relations":611,"tags":613,"workflow":614},1778699106670.4255,"k173sze6h2kdjhyfbveynf98t586n5wt",{"reviewCount":8},{"description":595,"installMethods":596,"name":598,"sourceUrl":599},"Use when the user wants to translate a repository README, make a repo multilingual, localize docs, add a language switcher, internationalize the README, or update localized README variants in a GitHub-style repository.",{"claudeCode":597},"xixu-me/skills","readme-i18n","https://github.com/xixu-me/skills",{"basePath":601,"githubOwner":602,"githubRepo":515,"locale":18,"slug":598,"type":247},"skills/readme-i18n","xixu-me",{"evaluate":604,"extract":610},{"promptVersionExtension":205,"promptVersionScoring":206,"score":547,"tags":605,"targetMarket":217,"tier":218},[606,607,608,214,609],"localization","documentation","internationalization","github",{"commitSha":278},{"repoId":612},"kd77r2vb42jmgam0qbr9f2c6kn86mebv",[607,609,608,606,214],{"evaluatedAt":615,"extractAt":616,"updatedAt":617},1778699174967,1778699106670,1778699384013,{"_creationTime":619,"_id":620,"community":621,"display":622,"identity":628,"providers":632,"relations":640,"tags":643,"workflow":644},1778691104676.0076,"k17bs66raa8vz3q5ysjnzhtvc586m7j3",{"reviewCount":8},{"description":623,"installMethods":624,"name":626,"sourceUrl":627},"Generate CSV, Markdown, and XLSX spreadsheets from structured tabular data.",{"claudeCode":625},"iterationlayer/skills","Spreadsheet Generation API","https://github.com/iterationlayer/skills",{"basePath":629,"githubOwner":630,"githubRepo":515,"locale":18,"slug":631,"type":247},"skills/sheet-generation-api","iterationlayer","sheet-generation-api",{"evaluate":633,"extract":639},{"promptVersionExtension":205,"promptVersionScoring":206,"score":547,"tags":634,"targetMarket":217,"tier":218},[635,636,637,214,638],"spreadsheet","csv","xlsx","data-generation",{"commitSha":278,"license":239},{"parentExtensionId":641,"repoId":642},"k1721s0xmp59902ybtpakrrffn86n10s","kd76p4g2qmtrkgx99cnab3683d86n4g8",[636,638,214,635,637],{"evaluatedAt":645,"extractAt":646,"updatedAt":645},1778693905213,1778691104676,{"_creationTime":648,"_id":649,"community":650,"display":651,"identity":657,"providers":662,"relations":670,"tags":672,"workflow":673},1778691799740.4802,"k173gtrv9a9ev95755pa3w820h86n7r2",{"reviewCount":8},{"description":652,"installMethods":653,"name":655,"sourceUrl":656},"Comprehensive markdown and Mermaid diagram writing skill. Use when creating any scientific document, report, analysis, or visualization. Establishes text-based diagrams as the default documentation standard with full style guides (markdown + mermaid), 24 diagram type references, and 9 document templates.",{"claudeCode":654},"K-Dense-AI/claude-scientific-skills","Markdown and Mermaid Writing","https://github.com/K-Dense-AI/claude-scientific-skills",{"basePath":658,"githubOwner":659,"githubRepo":660,"locale":18,"slug":661,"type":247},"scientific-skills/markdown-mermaid-writing","K-Dense-AI","claude-scientific-skills","markdown-mermaid-writing",{"evaluate":663,"extract":668},{"promptVersionExtension":205,"promptVersionScoring":206,"score":547,"tags":664,"targetMarket":217,"tier":218},[607,214,665,666,667],"mermaid","scientific-writing","diagramming",{"commitSha":278,"license":669},"Apache-2.0",{"repoId":671},"kd79rphh5gexy91xmpxc05h5mh86mm9r",[667,607,214,665,666],{"evaluatedAt":674,"extractAt":675,"updatedAt":674},1778693085426,1778691799740]