[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"extension-skill-Nagendhra-web-memory-bank-en":3,"guides-for-Nagendhra-web-memory-bank":391,"similar-k1723rcremvxvx59dy10cjc0f186n67k-en":392},{"_creationTime":4,"_id":5,"children":6,"community":7,"display":9,"evaluation":15,"identity":252,"isFallback":235,"parentExtension":256,"providers":307,"relations":311,"repo":312,"tags":388,"workflow":389},1778694389360.6567,"k1723rcremvxvx59dy10cjc0f186n67k",[],{"reviewCount":8},0,{"description":10,"installMethods":11,"name":13,"sourceUrl":14},"Token-efficient persistent memory system for Claude Code that saves ~67% tokens on session warm-up (verified with tiktoken). Layered architecture with progressive loading, compact encoding, branch-aware context, smart compression, session diffing, conflict detection, session continuation protocol, and recovery mode. Activates at session start (if MEMORY.md exists), on \"remember this\", \"pick up where we left off\", \"what were we doing\", \"wrap up\", \"save progress\", \"don't forget\", \"switch context\", \"hand off\", \"memory health\", \"save state\", \"continue where I left off\", \"context budget\", \"how much context left\", or any session start on a project with existing memory files. This skill solves two problems at once: Claude forgetting everything between sessions, AND sessions hitting context limits too fast. It replaces thousands of wasted re-explanation tokens with a compact, structured memory load that gives Claude full project context in under 2,000 tokens.\n",{"claudeCode":12},"Nagendhra-web/memory-bank","memory-bank","https://github.com/Nagendhra-web/memory-bank",{"_creationTime":16,"_id":17,"extensionId":5,"locale":18,"result":19,"trustSignals":233,"workflow":250},1778694438899.56,"kn7faqfhtej4z7s57w6qyatn0986nfr0","en",{"checks":20,"evaluatedAt":191,"extensionSummary":192,"features":193,"nonGoals":199,"practices":203,"prerequisites":208,"promptVersionExtension":211,"promptVersionScoring":212,"purpose":213,"rationale":214,"score":215,"summary":216,"tags":217,"targetMarket":226,"tier":227,"useCases":228},[21,26,29,32,36,39,43,46,50,54,58,61,64,68,72,76,78,81,84,87,91,95,98,102,105,108,111,114,117,120,124,127,131,135,139,142,145,148,152,155,158,161,164,167,170,174,178,181,184,188],{"category":22,"check":23,"severity":24,"summary":25},"Practical Utility","Problem relevance","pass","The description clearly states that the extension addresses Claude forgetting context between sessions and hitting context limits too quickly, which are concrete user problems.",{"category":22,"check":27,"severity":24,"summary":28},"Unique selling proposition","The extension offers significant value over basic prompting by providing a structured, token-efficient memory system with features like progressive loading, branch-aware context, and session continuation, which goes far beyond default LLM capabilities.",{"category":22,"check":30,"severity":24,"summary":31},"Production readiness","The skill is production-ready, offering a complete lifecycle for memory management, including load, save, compression, recovery, and session continuation, with clear documentation and examples.",{"category":33,"check":34,"severity":24,"summary":35},"Scope","Single responsibility principle","The extension focuses solely on managing Claude's session memory and context persistence, not bundling unrelated capabilities like code execution or external tool integrations.",{"category":33,"check":37,"severity":24,"summary":38},"Description quality","The displayed description accurately and concisely reflects the extension's core functionality, benefits, and activation triggers.",{"category":40,"check":41,"severity":24,"summary":42},"Invocation","Precise Purpose","The description clearly names the artifact (Claude's memory/context) and the user intent (saving context, extending session length) and details activation triggers and boundaries.",{"category":40,"check":44,"severity":24,"summary":45},"Concise Frontmatter","The frontmatter is dense and self-contained, accurately summarizing the core capability and providing precise trigger phrases within a reasonable character count.",{"category":47,"check":48,"severity":24,"summary":49},"Documentation","Concise Body","The SKILL.md body is concise, using separate reference files for deep dives and examples, keeping the core instructions focused.",{"category":51,"check":52,"severity":24,"summary":53},"Context","Progressive Disclosure","The SKILL.md outlines the flow and correctly links to separate reference files for detailed explanations of memory layers, compression, and branch awareness.",{"category":51,"check":55,"severity":56,"summary":57},"Forked exploration","not_applicable","The skill manages memory and context for Claude's responses, it does not involve deep code exploration or auditing that would necessitate `context: fork`.",{"category":22,"check":59,"severity":24,"summary":60},"Usage examples","Sufficient examples are provided, covering different project types and demonstrating the intended use and output of memory management.",{"category":22,"check":62,"severity":24,"summary":63},"Edge cases","The documentation addresses edge cases like memory corruption, stale memory, branch switching, and context limits with clear symptoms and recovery steps.",{"category":65,"check":66,"severity":56,"summary":67},"Code Execution","Tool Fallback","This skill operates within Claude's internal tool-use system and does not rely on external MCP servers or custom tools that would require fallbacks.",{"category":69,"check":70,"severity":24,"summary":71},"Safety","Halt on unexpected state","The documentation implies robust handling of state by emphasizing memory health checks, recovery modes, and structured diffing, which would halt on significant unexpected states.",{"category":73,"check":74,"severity":24,"summary":75},"Portability","Cross-skill coupling","The skill is self-contained and focuses on memory management; it does not implicitly rely on other skills and cross-references are explicit.",{"category":40,"check":44,"severity":24,"summary":77},"The frontmatter is concise, self-contained, and provides precise trigger phrases within a reasonable character count.",{"category":47,"check":79,"severity":56,"summary":80},"Configuration & parameter reference","There are no explicit configuration parameters or environment variables documented for this skill beyond the CLAUDE.md integration, which is optional.",{"category":33,"check":82,"severity":56,"summary":83},"Tool naming","This skill does not expose user-facing tools or commands; its functionality is triggered via natural language prompts and CLAUDE.md configuration.",{"category":33,"check":85,"severity":56,"summary":86},"Minimal I/O surface","The skill's interaction is through natural language prompts and CLAUDE.md configuration, not through defined tool parameter schemas or response shapes.",{"category":88,"check":89,"severity":24,"summary":90},"License","License usability","The extension is licensed under Apache-2.0, a permissive open-source license, clearly stated in the LICENSE file and README.",{"category":92,"check":93,"severity":24,"summary":94},"Maintenance","Commit recency","The last commit was on April 15, 2026, which is within the last 3 months, indicating active maintenance.",{"category":92,"check":96,"severity":56,"summary":97},"Dependency Management","The skill appears to have no third-party dependencies that require explicit management beyond the Python environment for auditing scripts.",{"category":99,"check":100,"severity":24,"summary":101},"Security","Secret Management","The skill explicitly warns against storing secrets in memory files and relies on Claude's secure handling of its own context, with no indication of secret leakage.",{"category":99,"check":103,"severity":24,"summary":104},"Injection","The skill focuses on structuring and presenting existing information, and there are no indications of it loading or executing untrusted third-party data as instructions.",{"category":99,"check":106,"severity":24,"summary":107},"Transitive Supply-Chain Grenades","The skill operates on local files and Claude's context, with no runtime downloads or remote code execution, thus avoiding supply-chain risks.",{"category":99,"check":109,"severity":24,"summary":110},"Sandbox Isolation","The skill's operations are confined to managing memory files and influencing Claude's internal context, with no file system writes or operations outside the project/user scope.",{"category":99,"check":112,"severity":24,"summary":113},"Sandbox escape primitives","There are no indications of detached process spawns, retry loops around denied calls, or other sandbox escape mechanisms in the provided code.",{"category":99,"check":115,"severity":24,"summary":116},"Data Exfiltration","The skill explicitly warns against storing secrets and focuses on project context, with no outbound calls to third parties for telemetry or data submission.",{"category":99,"check":118,"severity":24,"summary":119},"Hidden Text Tricks","The bundled content and documentation appear free of hidden steering tricks, invisible characters, or obfuscated instructions.",{"category":121,"check":122,"severity":24,"summary":123},"Hooks","Opaque code execution","The skill's logic is presented in readable Markdown and Python scripts, with no signs of obfuscation, base64 payloads, or runtime code fetching.",{"category":73,"check":125,"severity":24,"summary":126},"Structural Assumption","The skill correctly uses relative paths for memory files and assumes standard Git repositories, with clear guidance on CLAUDE.md placement.",{"category":128,"check":129,"severity":24,"summary":130},"Trust","Issues Attention","With 6 open and 0 closed issues in the last 90 days, the closure rate is N/A, but the low number of open issues suggests good engagement or a stable project.",{"category":132,"check":133,"severity":24,"summary":134},"Versioning","Release Management","The skill has a clear version (2.0.0) in the frontmatter and README, and install instructions reference a specific version, not 'main'.",{"category":136,"check":137,"severity":24,"summary":138},"Execution","Validation","The Python audit script uses basic file I/O validation, and the skill's core logic relies on Markdown parsing and Claude's internal prompt processing, which are implicitly validated.",{"category":99,"check":140,"severity":24,"summary":141},"Unguarded Destructive Operations","The skill is primarily read-only regarding project files; modifications are to memory files and CLAUDE.md, managed with clear user prompts and explicit user confirmation for actions like archival.",{"category":65,"check":143,"severity":24,"summary":144},"Error Handling","The audit script includes basic error handling for file access. The core skill logic relies on Claude's robust error handling for prompt processing and interaction.",{"category":65,"check":146,"severity":24,"summary":147},"Logging","The Python script provides audit logs for its own execution. The core skill's logging is managed implicitly by Claude's conversation transcript and the saving of memory files.",{"category":149,"check":150,"severity":24,"summary":151},"Compliance","GDPR","The skill explicitly warns against storing personal data or secrets, thus not operating on data that would require specific GDPR sanitization.",{"category":149,"check":153,"severity":24,"summary":154},"Target market","The extension is language-agnostic and operates on code and project context, with no regional or jurisdictional limitations, making it global.",{"category":73,"check":156,"severity":24,"summary":157},"Runtime stability","The skill is described in Markdown and uses standard Python for auditing, making it portable across common development environments without specific OS or shell assumptions.",{"category":47,"check":159,"severity":24,"summary":160},"README","The README is comprehensive, clearly states the extension's purpose, and effectively communicates its value proposition and features.",{"category":33,"check":162,"severity":56,"summary":163},"Tool surface size","This is a skill-based extension that interacts via natural language prompts and CLAUDE.md configuration, not through a list of exposed tools or commands.",{"category":40,"check":165,"severity":56,"summary":166},"Overlapping near-synonym tools","The extension interacts via natural language prompts and CLAUDE.md configuration, not through a command-line interface with multiple tools.",{"category":47,"check":168,"severity":24,"summary":169},"Phantom features","All advertised features, such as progressive loading, branch awareness, and compression, are well-documented and correspond to implemented logic in the SKILL.md and references.",{"category":171,"check":172,"severity":24,"summary":173},"Install","Installation instruction","Clear installation instructions are provided via `npx skills add` and a copy-pasteable CLAUDE.md snippet for auto-setup.",{"category":175,"check":176,"severity":24,"summary":177},"Errors","Actionable error messages","The documentation anticipates and addresses potential errors with clear troubleshooting steps and recovery procedures.",{"category":136,"check":179,"severity":24,"summary":180},"Pinned dependencies","The Python audit script lists `tiktoken` as a dependency, and while not explicitly pinned in the script itself, it's a common library. The main skill relies on Claude's runtime.",{"category":33,"check":182,"severity":24,"summary":183},"Dry-run preview","Operations like memory compression and archival are presented with clear previews and user confirmation prompts before execution.",{"category":185,"check":186,"severity":24,"summary":187},"Protocol","Idempotent retry & timeouts","The skill's operations are local file manipulations and prompt interactions, which do not involve remote calls requiring strict idempotency or timeouts beyond Claude's internal handling.",{"category":149,"check":189,"severity":24,"summary":190},"Telemetry opt-in","There is no mention of telemetry being collected or emitted by this skill.",1778694438788,"This skill provides a token-efficient, layered memory system for Claude Code, saving tokens on session warm-up and extending session length. It features progressive loading, branch-aware context, smart compression, session diffing, and recovery modes.",[194,195,196,197,198],"Token-efficient persistent memory system","Progressive loading and compact encoding","Branch-aware context and session diffing","Smart compression and archival","Session continuation protocol and recovery mode",[200,201,202],"Storing secrets or credentials within memory files","Replacing version control systems like Git for code changes","Executing arbitrary code or external commands beyond influencing Claude's context",[204,205,206,207],"Context Management","Session Persistence","Token Optimization","Developer Productivity",[209,210],"Claude Code environment","Git repository for branch-aware features","3.0.0","4.4.0","To ensure Claude Code retains context across sessions, eliminates token waste from re-explanation, and extends session duration by providing a structured, intelligent, and persistent memory system.","The extension is exceptionally well-documented and robust, addressing a critical user pain point with comprehensive features and clear instructions. Minor notes on dependency pinning for audit scripts and the lack of explicit retry mechanisms (due to its nature) prevent a perfect score.",99,"A highly polished and effective skill for managing Claude's session memory, significantly reducing token waste and improving context persistence.",[218,219,220,221,222,223,224,225],"memory","context","persistence","sessions","token-efficiency","branch-aware","compression","ai-assistant","global","verified",[229,230,231,232],"Reactivating a session without losing context or re-explaining project status","Working on projects with long-running tasks that exceed typical session limits","Collaborating with teammates by sharing and maintaining consistent project context","Onboarding new developers quickly with generated handoff documents",{"codeQuality":234,"collectedAt":236,"documentation":237,"maintenance":240,"security":246,"testCoverage":248},{"hasLockfile":235},false,1778694421559,{"descriptionLength":238,"readmeSize":239},966,18516,{"closedIssues90d":8,"forks":241,"hasChangelog":235,"manifestVersion":242,"openIssues90d":243,"pushedAt":244,"stars":245},10,"2.0.0",6,1776260095000,39,{"hasNpmPackage":235,"license":247,"smitheryVerified":235},"Apache-2.0",{"hasCi":249,"hasTests":235},true,{"updatedAt":251},1778694438899,{"basePath":253,"githubOwner":254,"githubRepo":13,"locale":18,"slug":13,"type":255},"skills/memory-bank","Nagendhra-web","skill",{"_creationTime":257,"_id":258,"community":259,"display":260,"identity":263,"parentExtension":265,"providers":296,"relations":302,"tags":303,"workflow":304},1778694389360.6565,"k17dbjbnpmrxhbszzw5sx6tsgs86n09d",{"reviewCount":8},{"description":261,"installMethods":262,"name":13,"sourceUrl":14},"Token-efficient persistent memory system with 3-tier architecture, branch-aware context, smart compression, session diffing, session continuation protocol, and recovery mode.",{"claudeCode":13},{"basePath":253,"githubOwner":254,"githubRepo":13,"locale":18,"slug":13,"type":264},"plugin",{"_creationTime":266,"_id":267,"community":268,"display":269,"identity":272,"providers":275,"relations":289,"tags":291,"workflow":292},1778694389360.6562,"k171ge78ys01r2ny6whcs987rs86n29s",{"reviewCount":8},{"description":270,"installMethods":271,"name":13,"sourceUrl":14},"Persistent memory for Claude Code that cuts token waste by 60-80% and makes sessions last 3-5x longer.",{"claudeCode":12},{"basePath":273,"githubOwner":254,"githubRepo":13,"locale":18,"slug":13,"type":274},"","marketplace",{"evaluate":276,"extract":283},{"promptVersionExtension":277,"promptVersionScoring":212,"score":278,"tags":279,"targetMarket":226,"tier":282},"3.1.0",75,[218,220,219,280,281],"efficiency","productivity","community",{"commitSha":284,"marketplace":285,"plugin":287},"HEAD",{"name":13,"pluginCount":286},1,{"mcpCount":8,"provider":288,"skillCount":8},"classify",{"repoId":290},"kd740s537z66ds6qge39y9dk3186nbp9",[219,280,218,220,281],{"evaluatedAt":293,"extractAt":294,"updatedAt":295},1778694403518,1778694389360,1778694473624,{"evaluate":297,"extract":301},{"promptVersionExtension":211,"promptVersionScoring":212,"score":298,"tags":299,"targetMarket":226,"tier":282},95,[218,219,220,221,222,223,224,300],"analytics",{"commitSha":284,"license":247},{"parentExtensionId":267,"repoId":290},[300,223,224,219,218,220,221,222],{"evaluatedAt":305,"extractAt":294,"updatedAt":306},1778694421287,1778694473855,{"evaluate":308,"extract":310},{"promptVersionExtension":211,"promptVersionScoring":212,"score":215,"tags":309,"targetMarket":226,"tier":227},[218,219,220,221,222,223,224,225],{"commitSha":284,"license":247},{"parentExtensionId":258,"repoId":290},{"_creationTime":313,"_id":290,"identity":314,"providers":315,"workflow":384},1778694384998.547,{"githubOwner":254,"githubRepo":13,"sourceUrl":14},{"classify":316,"discover":368,"github":371},{"commitSha":284,"extensions":317},[318,331,337],{"basePath":273,"description":270,"displayName":13,"installMethods":319,"rationale":320,"selectedPaths":321,"source":330,"sourceLanguage":18,"type":274},{"claudeCode":12},"marketplace.json at .claude-plugin/marketplace.json",[322,325,327],{"path":323,"priority":324},".claude-plugin/marketplace.json","mandatory",{"path":326,"priority":324},"README.md",{"path":328,"priority":329},"LICENSE","high","rule",{"basePath":253,"description":261,"displayName":13,"installMethods":332,"rationale":333,"selectedPaths":334,"source":330,"sourceLanguage":18,"type":264},{"claudeCode":13},"inline plugin source from marketplace.json at skills/memory-bank",[335],{"path":336,"priority":329},"SKILL.md",{"basePath":253,"description":10,"displayName":13,"installMethods":338,"rationale":339,"selectedPaths":340,"source":330,"sourceLanguage":18,"type":255},{"claudeCode":12},"SKILL.md frontmatter at skills/memory-bank/SKILL.md",[341,342,345,347,349,351,354,356,358,360,362,364,366],{"path":336,"priority":324},{"path":343,"priority":344},"examples/minimal.md","low",{"path":346,"priority":344},"examples/monorepo.md",{"path":348,"priority":344},"examples/solo-fullstack.md",{"path":350,"priority":344},"examples/team-backend.md",{"path":352,"priority":353},"references/advanced-patterns.md","medium",{"path":355,"priority":353},"references/branch-aware-memory.md",{"path":357,"priority":353},"references/claude-md-integration.md",{"path":359,"priority":353},"references/context-efficiency.md",{"path":361,"priority":353},"references/memory-layers.md",{"path":363,"priority":353},"references/session-diffing.md",{"path":365,"priority":353},"references/smart-compression.md",{"path":367,"priority":344},"scripts/token-audit.py",{"sources":369},[370],"manual",{"closedIssues90d":8,"description":372,"forks":241,"homepage":373,"license":247,"openIssues90d":243,"pushedAt":244,"readmeSize":239,"stars":245,"topics":374},"Persistent memory for Claude Code — 3-5x longer sessions, 60-80% fewer wasted tokens. Branch-aware, self-healing, token-efficient.","https://skills.sh/Nagendhra-web/memory-bank",[375,376,377,378,379,380,381,382,218,281,383,222],"agentskills","ai-agent","ai-skills","claude","claude-code","context-management","developer-tools","llm","skills-sh",{"classifiedAt":385,"discoverAt":386,"extractAt":387,"githubAt":387,"updatedAt":385},1778694389162,1778694384998,1778694387063,[225,223,224,219,218,220,221,222],{"evaluatedAt":251,"extractAt":294,"updatedAt":390},1778694474039,[],[393,421,448,477,507,537],{"_creationTime":394,"_id":395,"community":396,"display":397,"identity":403,"providers":406,"relations":413,"tags":416,"workflow":417},1778691765524.3735,"k175qp58x34qgtcj9b1tpxeern86mqvz",{"reviewCount":8},{"description":398,"installMethods":399,"name":401,"sourceUrl":402},"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":400},"juliusbrussee/caveman","caveman","https://github.com/juliusbrussee/caveman",{"basePath":404,"githubOwner":405,"githubRepo":401,"locale":18,"slug":401,"type":255},"skills/caveman","juliusbrussee",{"evaluate":407,"extract":411},{"promptVersionExtension":211,"promptVersionScoring":212,"score":215,"tags":408,"targetMarket":226,"tier":227},[224,382,222,409,410],"communication","prompt-engineering",{"commitSha":284,"license":412},"MIT",{"parentExtensionId":414,"repoId":415},"k170g496f5ze5hv8xa2e7b91ex86mz9m","kd7890f0sfy0ymf1pakpx7vqch86nmew",[409,224,382,410,222],{"evaluatedAt":418,"extractAt":419,"updatedAt":420},1778691864047,1778691765524,1778692174530,{"_creationTime":422,"_id":423,"community":424,"display":425,"identity":431,"providers":436,"relations":441,"tags":443,"workflow":444},1778698056313.1526,"k1790kka1gerjkd8t2y1bfct6586ne6s",{"reviewCount":8},{"description":426,"installMethods":427,"name":429,"sourceUrl":430},"Search claude-mem's persistent cross-session memory database. Use when user asks \"did we already solve this?\", \"how did we do X last time?\", or needs work from previous sessions.",{"claudeCode":428},"thedotmack/claude-mem","Claude-Mem","https://github.com/thedotmack/claude-mem",{"basePath":432,"githubOwner":433,"githubRepo":434,"locale":18,"slug":435,"type":255},"plugin/skills/mem-search","thedotmack","claude-mem","mem-search",{"evaluate":437,"extract":440},{"promptVersionExtension":211,"promptVersionScoring":212,"score":215,"tags":438,"targetMarket":226,"tier":227},[218,439,219,220,379],"search",{"commitSha":284,"license":247},{"repoId":442},"kd70jnxgm695az2wtf37zbqdj986mp7k",[379,219,218,220,439],{"evaluatedAt":445,"extractAt":446,"updatedAt":447},1778698211608,1778698056313,1778698443619,{"_creationTime":449,"_id":450,"community":451,"display":452,"identity":458,"providers":462,"relations":469,"tags":472,"workflow":473},1778697873170.672,"k172m47x5hg2kpyn0bmnjv1nhh86nbfh",{"reviewCount":8},{"description":453,"installMethods":454,"name":456,"sourceUrl":457},"Save important project knowledge to memory. Use when user wants to preserve architectural decisions, significant bug fixes, design patterns, or important implementation details for team reference.",{"claudeCode":455},"supermemoryai/claude-supermemory","super-save","https://github.com/supermemoryai/claude-supermemory",{"basePath":459,"githubOwner":460,"githubRepo":461,"locale":18,"slug":456,"type":255},"plugin/skills/super-save","supermemoryai","claude-supermemory",{"evaluate":463,"extract":468},{"promptVersionExtension":211,"promptVersionScoring":212,"score":464,"tags":465,"targetMarket":226,"tier":227},98,[218,466,467,225,220],"knowledge-management","project-documentation",{"commitSha":284},{"parentExtensionId":470,"repoId":471},"k177gnvdhpev80y7wbcqhfv7e186mkab","kd7d4aasa10djz2xs8fmv0w16186n8rx",[225,466,218,220,467],{"evaluatedAt":474,"extractAt":475,"updatedAt":476},1778697914987,1778697873170,1778697967427,{"_creationTime":478,"_id":479,"community":480,"display":481,"identity":487,"providers":491,"relations":500,"tags":503,"workflow":504},1778686640222.7917,"k17549bwd8rq0h0wq4cvav0fpn86nda9",{"reviewCount":8},{"description":482,"installMethods":483,"name":485,"sourceUrl":486},"Explore requirements and approaches through collaborative dialogue, then write a right-sized requirements document. Use when the user says \"let's brainstorm\", \"what should we build\", or \"help me think through X\", presents a vague or ambitious feature request, or seems unsure about scope or direction -- even without explicitly asking to brainstorm.",{"claudeCode":484},"EveryInc/compound-engineering-plugin","ce-brainstorm","https://github.com/EveryInc/compound-engineering-plugin",{"basePath":488,"githubOwner":489,"githubRepo":490,"locale":18,"slug":485,"type":255},"plugins/compound-engineering/skills/ce-brainstorm","EveryInc","compound-engineering-plugin",{"evaluate":492,"extract":499},{"promptVersionExtension":211,"promptVersionScoring":212,"score":493,"tags":494,"targetMarket":226,"tier":227},100,[495,496,497,498,225],"documentation","requirements","collaboration","planning",{"commitSha":284},{"parentExtensionId":501,"repoId":502},"k17d893df4em0e3pn8f55h1dxn86n09e","kd7e40my1b5g70tg0f60qg85ch86nn08",[225,497,495,498,496],{"evaluatedAt":505,"extractAt":506,"updatedAt":505},1778686797367,1778686640222,{"_creationTime":508,"_id":509,"community":510,"display":511,"identity":517,"providers":522,"relations":530,"tags":533,"workflow":534},1778694269038.6682,"k1752cypc448mke749yjbkc65186mg6f",{"reviewCount":8},{"description":512,"installMethods":513,"name":515,"sourceUrl":516},"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":514},"muratcankoylan/Agent-Skills-for-Context-Engineering","Context Compression","https://github.com/muratcankoylan/Agent-Skills-for-Context-Engineering",{"basePath":518,"githubOwner":519,"githubRepo":520,"locale":18,"slug":521,"type":255},"skills/context-compression","muratcankoylan","Agent-Skills-for-Context-Engineering","context-compression",{"evaluate":523,"extract":529},{"promptVersionExtension":211,"promptVersionScoring":212,"score":493,"tags":524,"targetMarket":226,"tier":227},[525,382,526,527,224,528],"context-engineering","agent","summarization","evaluation",{"commitSha":284,"license":412},{"parentExtensionId":531,"repoId":532},"k1754dy3wbsv2a5gr1a983zzs586njca","kd7f12maf5nxmx5xttjx7scfnx86m1tv",[526,224,525,528,382,527],{"evaluatedAt":535,"extractAt":536,"updatedAt":535},1778694410149,1778694269038,{"_creationTime":538,"_id":539,"community":540,"display":541,"identity":547,"providers":551,"relations":559,"tags":562,"workflow":563},1778691497595.9666,"k170aa7za6n5wvxtwx7xj8dc9d86mx9e",{"reviewCount":8},{"description":542,"installMethods":543,"name":545,"sourceUrl":546},"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":544},"jimliu/baoyu-skills","baoyu-compress-image","https://github.com/jimliu/baoyu-skills",{"basePath":548,"githubOwner":549,"githubRepo":550,"locale":18,"slug":545,"type":255},"skills/baoyu-compress-image","jimliu","baoyu-skills",{"evaluate":552,"extract":558},{"promptVersionExtension":211,"promptVersionScoring":212,"score":493,"tags":553,"targetMarket":226,"tier":227},[554,224,555,556,557],"image","webp","png","optimization",{"commitSha":284},{"parentExtensionId":560,"repoId":561},"k172q4hxm2acmg5ree5aeemx1x86m6hx","kd718jn448p61w289hzf15tecd86n3a4",[224,554,557,556,555],{"evaluatedAt":564,"extractAt":565,"updatedAt":564},1778691619519,1778691497596]