[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"extension-marketplace-Anjos2-recursive-research-en":3,"guides-for-Anjos2-recursive-research":322,"similar-k171vfst8y7c4bkw07sx66bgb986n8e0-en":323},{"_creationTime":4,"_id":5,"children":6,"community":49,"display":50,"evaluation":54,"identity":257,"isFallback":244,"parentExtension":260,"providers":261,"relations":267,"repo":268,"tags":319,"workflow":320},1778675291492.648,"k171vfst8y7c4bkw07sx66bgb986n8e0",[7],{"_creationTime":8,"_id":9,"community":10,"display":12,"identity":18,"providers":23,"relations":42,"tags":44,"workflow":45},1778675291492.6482,"k178rg6vs234c409g28ca6arbd86n5a1",{"reviewCount":11},0,{"description":13,"installMethods":14,"name":16,"sourceUrl":17},"Recursive research up to PhD level across any domain (science, tech, business, arts, humanities). Source tiering, WDM + Munger inversion for autonomous decisions, disk checkpointing to survive context compaction.",{"claudeCode":15},"recursive-research","Recursive Research","https://github.com/Anjos2/recursive-research",{"basePath":19,"githubOwner":20,"githubRepo":15,"locale":21,"slug":15,"type":22},"plugins/recursive-research","Anjos2","en","plugin",{"evaluate":24,"extract":36},{"promptVersionExtension":25,"promptVersionScoring":26,"score":27,"tags":28,"targetMarket":34,"tier":35},"3.0.0","4.4.0",95,[29,30,31,32,33],"research","knowledge-management","ai-agent","information-retrieval","claude-code","global","verified",{"commitSha":37,"license":38,"plugin":39},"HEAD","MIT",{"mcpCount":11,"provider":40,"skillCount":41},"classify",1,{"parentExtensionId":5,"repoId":43},"kd753d1f20n4nwaapq6yp8vhd186nzxp",[31,33,32,30,29],{"evaluatedAt":46,"extractAt":47,"updatedAt":48},1778675336950,1778675291492,1778675391850,{"reviewCount":11},{"description":51,"installMethods":52,"name":16,"sourceUrl":17},"Recursive research skill/plugin for Claude Code by Joseph Huayhualla (@Anjos2)",{"claudeCode":53},"Anjos2/recursive-research",{"_creationTime":55,"_id":56,"extensionId":5,"locale":21,"result":57,"trustSignals":242,"workflow":255},1778675309479.3374,"kn7190gde1052gpkxh7mgjcfs186n09t",{"checks":58,"evaluatedAt":205,"extensionSummary":206,"features":207,"nonGoals":213,"practices":218,"prerequisites":222,"promptVersionExtension":223,"promptVersionScoring":26,"purpose":224,"rationale":225,"score":27,"summary":226,"tags":227,"targetMarket":34,"tier":35,"useCases":229,"workflow":234},[59,64,67,70,74,77,82,87,90,93,97,101,104,108,111,114,117,120,123,126,130,134,138,142,146,149,152,155,159,162,165,168,171,174,177,181,185,189,192,196,199,202],{"category":60,"check":61,"severity":62,"summary":63},"Practical Utility","Problem relevance","pass","The description clearly states the problem of performing recursive research on any topic up to PhD level, identifying specific use cases like science, tech, business, arts, and humanities.",{"category":60,"check":65,"severity":62,"summary":66},"Unique selling proposition","The extension offers significant value over default LLM behavior by implementing features like source tiering, auto-regulation, disk checkpointing, WDM + Munger inversion for autonomous decisions, and a PhD fitness function, which are not standard LLM capabilities.",{"category":60,"check":68,"severity":62,"summary":69},"Production readiness","The extension provides a complete lifecycle for research, from seed input to consolidated knowledge and decision-making, including user interaction and persistence via disk checkpoints, making it production-ready.",{"category":71,"check":72,"severity":62,"summary":73},"Scope","Single responsibility principle","The extension focuses on recursive research, encompassing related functionalities like source tiering, decision-making, and knowledge consolidation within a single, coherent workflow.",{"category":71,"check":75,"severity":62,"summary":76},"Description quality","The displayed description accurately and concisely reflects the extension's core functionality as a recursive research plugin/skill.",{"category":78,"check":79,"severity":80,"summary":81},"Invocation","Scoped tools","not_applicable","This is a marketplace extension and does not expose tools directly; evaluation of tools would apply to hosted items if this were a skill or plugin.",{"category":83,"check":84,"severity":85,"summary":86},"Documentation","Configuration & parameter reference","info","The README details interactive prompts and cycle caps, and `--resume` and `--list` flags are mentioned, but a formal schema or explicit default values for all parameters are not provided.",{"category":71,"check":88,"severity":80,"summary":89},"Tool naming","As this is a marketplace extension, it does not expose user-facing tools directly for evaluation of their naming.",{"category":71,"check":91,"severity":80,"summary":92},"Minimal I/O surface","This marketplace extension does not have tools with parameter schemas or response shapes to evaluate.",{"category":94,"check":95,"severity":62,"summary":96},"License","License usability","The LICENSE file clearly states the MIT license, which is a permissive open-source license, and this is also referenced in the README and marketplace manifest.",{"category":98,"check":99,"severity":62,"summary":100},"Maintenance","Commit recency","The last commit was on April 22, 2026, which is within the last 3 months, indicating recent maintenance.",{"category":98,"check":102,"severity":80,"summary":103},"Dependency Management","The extension does not appear to rely on external third-party dependencies that would require specific management or vulnerability checks.",{"category":105,"check":106,"severity":62,"summary":107},"Security","Secret Management","The README states 'recursive-research does not collect any user data. Everything runs locally on your machine.' and there are no indications of secret handling that would leak into logs.",{"category":105,"check":109,"severity":62,"summary":110},"Injection","The extension's design, focusing on local processing and documented inputs/outputs, suggests that loaded data is treated as information rather than executable instructions. No runtime downloads of external content are described.",{"category":105,"check":112,"severity":62,"summary":113},"Transitive Supply-Chain Grenades","The extension's implementation appears to keep all content within the bundle and local filesystem, with no runtime downloads or execution of external code described.",{"category":105,"check":115,"severity":62,"summary":116},"Sandbox Isolation","The extension operates locally within the user's project directory ('memoria/investigaciones/'), creating files only within this scope and not modifying external system files or configurations.",{"category":105,"check":118,"severity":62,"summary":119},"Sandbox escape primitives","No detached process spawns or retry loops around denied tool calls are evident from the provided documentation.",{"category":105,"check":121,"severity":62,"summary":122},"Data Exfiltration","The extension explicitly states it does not collect user data and runs locally, with no outbound calls described for telemetry or submission of confidential information.",{"category":105,"check":124,"severity":62,"summary":125},"Hidden Text Tricks","The provided README and other documentation files appear to be free of hidden text tricks, invisible Unicode characters, or other obfuscation techniques.",{"category":127,"check":128,"severity":62,"summary":129},"Hooks","Opaque code execution","The provided documentation describes a skill implemented in markdown, with no indications of obfuscated code, base64 payloads, or runtime script fetching.",{"category":131,"check":132,"severity":62,"summary":133},"Portability","Structural Assumption","The extension explicitly creates a 'memoria/' directory within the active project if it doesn't exist, indicating awareness of potential layout differences and handling them gracefully.",{"category":135,"check":136,"severity":62,"summary":137},"Trust","Issues Attention","There are 0 open and 0 closed issues in the last 90 days, indicating minimal activity but no signs of neglect.",{"category":139,"check":140,"severity":62,"summary":141},"Versioning","Release Management","The extension has a declared version (2.2.0) in the README and marketplace manifest, and the commit history indicates recent updates.",{"category":143,"check":144,"severity":85,"summary":145},"Code Execution","Validation","The extension relies on interactive prompts and user input. While parameters like cycle caps are mentioned, there's no explicit mention of schema validation libraries for these inputs or for structured output.",{"category":105,"check":147,"severity":62,"summary":148},"Unguarded Destructive Operations","The primary operations involve file writes for checkpoints and research data, which are confined to the user's project directory and are not inherently destructive without user interaction.",{"category":143,"check":150,"severity":85,"summary":151},"Error Handling","The README mentions that the skill guides the user interactively and asks questions, implying some error handling for user input. However, details on structured error reporting or fail-closed behavior for internal operations are not provided.",{"category":143,"check":153,"severity":80,"summary":154},"Logging","The extension's documentation indicates that all operations are local and private, and there is no mention of explicit audit logging to a local file. This is acceptable given the privacy focus.",{"category":156,"check":157,"severity":62,"summary":158},"Compliance","GDPR","The extension explicitly states that it does not collect any user data and runs locally, thus avoiding any GDPR concerns related to personal data.",{"category":156,"check":160,"severity":62,"summary":161},"Target market","The extension's functionality is universally applicable across all research domains and is designed to run locally, making it suitable for a global target market.",{"category":131,"check":163,"severity":62,"summary":164},"Runtime stability","The extension is designed to run locally and interactively, with no specific OS or shell dependencies mentioned beyond standard file system operations and Claude Code's environment.",{"category":83,"check":166,"severity":62,"summary":167},"README","The README is comprehensive, clearly states the extension's purpose, features, installation, and usage, and is well-structured.",{"category":71,"check":169,"severity":80,"summary":170},"Tool surface size","This is a marketplace extension; tool surface size is not applicable.",{"category":78,"check":172,"severity":80,"summary":173},"Overlapping near-synonym tools","As a marketplace extension, there are no direct tools to evaluate for overlapping synonyms.",{"category":83,"check":175,"severity":62,"summary":176},"Phantom features","All features described in the README, such as source tiering, checkpointing, and WDM + Munger inversion, appear to be implemented based on the detailed explanation of the research process.",{"category":178,"check":179,"severity":62,"summary":180},"Install","Installation instruction","The README provides multiple clear, copy-pasteable installation options (marketplace, plugin directory, standalone skill) with example invocations.",{"category":182,"check":183,"severity":85,"summary":184},"Errors","Actionable error messages","While the extension guides the user interactively, specific details about the format and content of error messages, including remediation steps, are not detailed in the documentation.",{"category":186,"check":187,"severity":80,"summary":188},"Execution","Pinned dependencies","The extension appears to be a self-contained skill/plugin written in markdown and does not list external code dependencies that would require pinning.",{"category":71,"check":190,"severity":80,"summary":191},"Dry-run preview","The extension is primarily analytical and knowledge-gathering, with file writes for local storage. A specific dry-run mode for these operations is not described or evidently necessary.",{"category":193,"check":194,"severity":80,"summary":195},"Protocol","Idempotent retry & timeouts","The extension operates locally and interactively, without remote calls or state-changing operations that would necessitate idempotency or timeouts.",{"category":156,"check":197,"severity":62,"summary":198},"Telemetry opt-in","The extension explicitly states that it does not collect any user data, meaning no telemetry is emitted, fulfilling the opt-in requirement by default.",{"category":71,"check":200,"severity":62,"summary":201},"Theme declaration","The marketplace README clearly declares a curation theme focused on recursive research tools, including specific criteria like source tiering and autonomous decision-making.",{"category":139,"check":203,"severity":62,"summary":204},"Per-entry version metadata","The marketplace manifest specifies a version (2.2.0) for the 'recursive-research' plugin entry, providing explicit version metadata.",1778675309124,"This is a Claude Code marketplace entry for a recursive research plugin/skill that operates locally, performing in-depth research across various domains by tiering sources, employing weighted decision matrices and Munger inversion, and saving progress via disk checkpoints. It aims to achieve 'PhD level' research criteria.",[208,209,210,211,212],"Recursive research across any domain","Source tiering and rejection of low-quality sources","Disk checkpointing for resilience","Autonomous decision-making via WDM and Munger inversion","PhD fitness function for research completion criteria",[214,215,216,217],"Collecting or transmitting any user data","Performing actions outside the user's specified project directory","Relying on external scripts or opaque code execution","Acting as a general-purpose web scraper without a research objective",[219,220,221],"Research methodology","Information synthesis","Autonomous decision-making",[],"3.1.0","To provide an advanced, autonomous, and locally-run recursive research tool that can tackle complex topics and deliver high-quality, curated knowledge, surviving context limitations.","The extension has no critical or warning findings. It excels in security, privacy, documentation, and production readiness. Minor info findings for parameter validation and error message detail do not significantly impact its verified status.","A highly polished and production-ready recursive research skill/plugin with strong security and privacy guarantees.",[31,228,30,15,33],"research-tool",[230,231,232,233],"Conducting in-depth academic research on complex topics","Gathering and synthesizing information for scientific papers or dissertations","Analyzing business trends or technological advancements with high detail","Exploring historical or philosophical subjects exhaustively",[235,236,237,238,239,240,241],"User provides research seed, mode, paths, sources, and cycle cap.","Skill identifies seed threads using WDM + Munger inversion.","Skill detects and prioritizes available MCPs (e.g., Firecrawl, WebFetch).","Skill iterates in auto-regulated cycles, investigating least-covered threads.","Sources are tiered, and disk checkpoints are saved each cycle.","Skill closes when PhD fitness criteria are met or cycle cap is reached.","User is prompted to continue research if applicable.",{"codeQuality":243,"collectedAt":245,"documentation":246,"maintenance":249,"security":253,"testCoverage":254},{"hasLockfile":244},false,1778675293244,{"descriptionLength":247,"readmeSize":248},78,10252,{"closedIssues90d":11,"forks":11,"hasChangelog":244,"manifestVersion":250,"openIssues90d":11,"pushedAt":251,"stars":252},"2.2.0",1776902259000,7,{"hasNpmPackage":244,"license":38,"smitheryVerified":244},{"hasCi":244,"hasTests":244},{"updatedAt":256},1778675309479,{"basePath":258,"githubOwner":20,"githubRepo":15,"locale":21,"slug":15,"type":259},"","marketplace",null,{"evaluate":262,"extract":264},{"promptVersionExtension":223,"promptVersionScoring":26,"score":27,"tags":263,"targetMarket":34,"tier":35},[31,228,30,15,33],{"commitSha":37,"license":38,"marketplace":265,"plugin":266},{"name":15,"pluginCount":41},{"mcpCount":11,"provider":40,"skillCount":11},{"repoId":43},{"_creationTime":269,"_id":43,"identity":270,"providers":271,"workflow":315},1778675287813.5454,{"githubOwner":20,"githubRepo":15,"sourceUrl":17},{"classify":272,"discover":306,"github":309},{"commitSha":37,"extensions":273},[274,287,296],{"basePath":258,"description":51,"displayName":15,"installMethods":275,"rationale":276,"selectedPaths":277,"source":286,"sourceLanguage":21,"type":259},{"claudeCode":53},"marketplace.json at .claude-plugin/marketplace.json",[278,281,283],{"path":279,"priority":280},".claude-plugin/marketplace.json","mandatory",{"path":282,"priority":280},"README.md",{"path":284,"priority":285},"LICENSE","high","rule",{"basePath":19,"description":13,"displayName":15,"installMethods":288,"license":38,"rationale":289,"selectedPaths":290,"source":286,"sourceLanguage":21,"type":22},{"claudeCode":15},"plugin manifest at plugins/recursive-research/.claude-plugin/plugin.json",[291,293],{"path":292,"priority":280},".claude-plugin/plugin.json",{"path":294,"priority":295},"skills/recursive-research/SKILL.md","medium",{"basePath":297,"description":298,"displayName":15,"installMethods":299,"rationale":300,"selectedPaths":301,"source":286,"sourceLanguage":304,"type":305},"plugins/recursive-research/skills/recursive-research","Investigación recursiva profunda con loop auto-regulado hasta nivel PhD. Aplicable a cualquier dominio (ciencia, tecnología, negocio, arte, humanidades). Usa WDM + Inversión Munger para decisiones autónomas, tiering de fuentes confiables, y checkpointing a disco para sobrevivir límites de contexto.",{"claudeCode":53},"SKILL.md frontmatter at plugins/recursive-research/skills/recursive-research/SKILL.md",[302],{"path":303,"priority":280},"SKILL.md","es","skill",{"sources":307},[308],"manual",{"closedIssues90d":11,"description":310,"forks":11,"license":38,"openIssues90d":11,"pushedAt":251,"readmeSize":248,"stars":252,"topics":311},"Claude Code skill for recursive research up to PhD level across any domain. Source tiering, WDM + Munger inversion for autonomous decisions, and disk checkpointing to survive context compaction.",[31,33,312,30,313,228,15,314],"claude-code-skill","mental-models","weighted-decision-matrix",{"classifiedAt":316,"discoverAt":317,"extractAt":318,"githubAt":318,"updatedAt":316},1778675291330,1778675287813,1778675289690,[31,33,30,15,228],{"evaluatedAt":256,"extractAt":47,"updatedAt":321},1778675391455,[],[324,354,384,416,445],{"_creationTime":325,"_id":326,"community":327,"display":328,"identity":334,"providers":336,"relations":347,"tags":349,"workflow":350},1778696712851.9736,"k1730g2nnwga3htdfycmc8h1gd86m3cs",{"reviewCount":11},{"description":329,"installMethods":330,"name":332,"sourceUrl":333},"2x the effective context with smart weight-loss for Claude Code — prune bloated sessions, protect agent teams from compaction, monitor token usage with MCP tools",{"claudeCode":331},"Ruya-AI/cozempic","cozempic","https://github.com/Ruya-AI/cozempic",{"basePath":258,"githubOwner":335,"githubRepo":332,"locale":21,"slug":332,"type":259},"Ruya-AI",{"evaluate":337,"extract":344},{"promptVersionExtension":223,"promptVersionScoring":26,"score":338,"tags":339,"targetMarket":34,"tier":35},100,[340,341,33,342,343],"context","optimization","session-management","ai-productivity",{"commitSha":37,"marketplace":345,"plugin":346},{"name":332,"pluginCount":41},{"mcpCount":11,"provider":40,"skillCount":11},{"repoId":348},"kd79d77qmyh3826dwhk7ynx9xd86nmqm",[343,33,340,341,342],{"evaluatedAt":351,"extractAt":352,"updatedAt":353},1778696729067,1778696712852,1778696924064,{"_creationTime":355,"_id":356,"community":357,"display":358,"identity":364,"providers":369,"relations":378,"tags":380,"workflow":381},1778691765524.3726,"k17c1ndsb7hncr0k5vadc69rnx86njvh",{"reviewCount":11},{"description":359,"installMethods":360,"name":362,"sourceUrl":363},"🪨 why use many token when few token do trick — Claude Code skill that cuts 65% of tokens by talking like caveman",{"claudeCode":361},"juliusbrussee/caveman","caveman-repo","https://github.com/juliusbrussee/caveman",{"basePath":365,"githubOwner":366,"githubRepo":367,"locale":21,"slug":368,"type":259},".agents/plugins","juliusbrussee","caveman","plugins",{"evaluate":370,"extract":376},{"promptVersionExtension":223,"promptVersionScoring":26,"score":338,"tags":371,"targetMarket":34,"tier":35},[372,373,374,33,375],"token-compression","llm-efficiency","cli","prompt-engineering",{"commitSha":37,"marketplace":377},{"name":362,"pluginCount":11},{"repoId":379},"kd7890f0sfy0ymf1pakpx7vqch86nmew",[33,374,373,375,372],{"evaluatedAt":382,"extractAt":383,"updatedAt":382},1778691782274,1778691765524,{"_creationTime":385,"_id":386,"community":387,"display":388,"identity":394,"providers":397,"relations":410,"tags":412,"workflow":413},1778683562157.875,"k174pnm5ch9ab6fr1etef2f2b586m74b",{"reviewCount":11},{"description":389,"installMethods":390,"name":392,"sourceUrl":393},"Persistent memory and cognitive profiling plugins for Claude Code",{"claudeCode":391},"cdeust/Cortex","cortex-plugins","https://github.com/cdeust/Cortex",{"basePath":258,"githubOwner":395,"githubRepo":396,"locale":21,"slug":396,"type":259},"cdeust","Cortex",{"evaluate":398,"extract":407},{"promptVersionExtension":223,"promptVersionScoring":26,"score":338,"tags":399,"targetMarket":34,"tier":35},[400,401,402,33,403,404,405,406],"memory","cognitive-profiling","mcp","knowledge-graph","codebase-analysis","postgresql","pgvector",{"commitSha":37,"marketplace":408,"plugin":409},{"name":392,"pluginCount":41},{"mcpCount":11,"provider":40,"skillCount":11},{"repoId":411},"kd79gxpemvkr09a7zsb3h8kmah86nvgf",[33,404,401,403,402,400,406,405],{"evaluatedAt":414,"extractAt":415,"updatedAt":414},1778683583007,1778683562157,{"_creationTime":417,"_id":418,"community":419,"display":420,"identity":426,"providers":428,"relations":438,"tags":440,"workflow":441},1778698388090.9072,"k171hvn0e2t3jzczxm3vzspj3186mgj7",{"reviewCount":11},{"description":421,"installMethods":422,"name":424,"sourceUrl":425},"Browser automation for AI agents",{"claudeCode":423},"vercel-labs/agent-browser","agent-browser","https://github.com/vercel-labs/agent-browser",{"basePath":258,"githubOwner":427,"githubRepo":424,"locale":21,"slug":424,"type":259},"vercel-labs",{"evaluate":429,"extract":435},{"promptVersionExtension":223,"promptVersionScoring":26,"score":430,"tags":431,"targetMarket":34,"tier":35},99,[432,374,433,434,31],"browser-automation","webdriver","testing",{"commitSha":37,"marketplace":436,"plugin":437},{"name":424,"pluginCount":41},{"mcpCount":11,"provider":40,"skillCount":11},{"repoId":439},"kd74a6sfc26xa0ec2tpntj8r2986mn0a",[31,432,374,434,433],{"evaluatedAt":442,"extractAt":443,"updatedAt":444},1778698406247,1778698388091,1778698675865,{"_creationTime":446,"_id":447,"community":448,"display":449,"identity":455,"providers":458,"relations":468,"tags":470,"workflow":471},1778696595410.5613,"k1741gyqaf0tj1ffq65gjpsm6x86mfhx",{"reviewCount":11},{"description":450,"installMethods":451,"name":453,"sourceUrl":454},"Battle-tested Claude Code workflows from power users",{"claudeCode":452},"rohitg00/pro-workflow","Pro Workflow","https://github.com/rohitg00/pro-workflow",{"basePath":258,"githubOwner":456,"githubRepo":457,"locale":21,"slug":457,"type":259},"rohitg00","pro-workflow",{"evaluate":459,"extract":465},{"promptVersionExtension":223,"promptVersionScoring":26,"score":430,"tags":460,"targetMarket":34,"tier":35},[461,462,463,464,30],"workflow","productivity","ai-assistance","developer-tools",{"commitSha":37,"license":38,"marketplace":466,"plugin":467},{"name":457,"pluginCount":41},{"mcpCount":11,"provider":40,"skillCount":11},{"repoId":469},"kd7am4e918eq98hrd9s31jm4vs86nn0b",[463,464,30,462,461],{"evaluatedAt":472,"extractAt":473,"updatedAt":472},1778696611346,1778696595410]