Context compression for URLs, files, images, PDFs, git diffs, PRs, and logs via caveman-mcp. Two-pass compression (mechanical filler removal + classical Chinese ultra-compression) preserves code, identifiers, and numbers while slashing tokens. Skills cover URL fetching with smart routing (YouTube/GitHub/arXiv/HN/Reddit/RSS), local file ingestion (PDF/PPT/XLSX/audio), image OCR/description, git diff/PR summarization, log triage with deduplication, raw text compression, and a multi-source research workflow.
Plugin caveman from marketplace standardbeagle-tools in standardbeagle/standardbeagle-tools. Commands read from the repo's manifest — 4 days ago.
caveman-condense-fileCompress local files via caveman-mcp — PDF/slides/spreadsheet/audio/text handlers. Use instead of Read when file too large or only gist matters. 以caveman壓縮本地文件。 Use when: ingest large PDF, summarize slide deck, scan spreadsheet, transcribe audio, condense long doc, prep file for downstream synthesis
caveman-condense-gitCompress git diffs, logs, blame output, and GitHub PR URLs via caveman-mcp. Use for PR review, change summarization, blame archaeology when raw diff exceeds context budget. 以caveman壓縮git diff/log/blame及GitHub PR URL. Use when: review large PR, summarize branch diff, archaeology git blame, ingest PR
caveman-condense-imageExtract description, OCR text, structural summary from images via caveman-mcp condense_file. For screenshots/diagrams/charts/photos when token budget rules out raw image. 以caveman提取圖像描述/OCR/結構。 Use when: ingest screenshot, summarize diagram, extract chart data, photo OCR, archive image as text, prep
caveman-condense-logCompress error logs + stack traces via caveman-mcp with language detection + dedup. For log triage, post-incident review, CI failure analysis when raw logs too long. 以caveman壓縮日誌與堆棧。 Use when: triage CI failure log, summarize app error output, dedupe stack traces, post-incident analysis, prep log fo
caveman-condense-textCompress raw text via caveman-mcp two-pass (filler removal + classical Chinese ultra). Code/identifiers/numbers preserved. 以caveman壓縮任意文本。 Use when: compress chat transcript, condense prose, prep meeting notes, summarize research draft, reduce memory entry size, ad-hoc inline compression
caveman-condense-urlFetch + compress any URL via caveman-mcp with smart routing for YouTube/GitHub/arXiv/HN/Reddit/RSS. Replaces WebFetch when token budget tight or content cited later. 以caveman取URL並壓縮。 Use when: research URLs, summarize articles, YouTube transcripts, ingest arXiv papers, scan HN/Reddit, follow RSS, pr
caveman-recommenderRoutes caveman / context-compression intent to right manual skill — sole auto gateway; others manual. caveman 唯一自動網關。 Use when: compress file/PDF/slides/spreadsheet/audio, condense git diff/log or PR URL, OCR/summarize image/screenshot/diagram, compress error logs/stack traces/notes, fetch+compress
caveman-research-workflowMulti-source research recipe via caveman-mcp — gather URLs/PDFs/PRs in parallel, compress each, synthesize with provenance. Pairs with knowledge-hygiene for conflict detection. 多源研究：並行採集、壓縮、含出處合成。 Use when: research topic across sources, prep literature review, ingest reading list, build evidence pa
caveman-setup-caveman\"Set up caveman-mcp server with SLOP management for context compression. 以SLOP管理配置caveman-mcp服務器以行上下文壓縮. Use when: install caveman, setup context compression, configure caveman-mcp, first time caveman setup, register condense tools\"
Browser superpowers for AI coding agents: dev-server and reverse proxy with traffic logging, a deduped priority error inbox, 50+ frontend diagnostic functions, accessibility/performance/security/SEO/responsive/API/loading audits, visual-regression snapshots, replay testing, tunnels, and sketch/design modes
36 skills · 4 agents · hooks
Dart task management with adversarial cooperation loops and full dart-query MCP schema parity - DartQL batch operations, advanced filtering, task relationships, documents, time tracking, and automated task execution through quality verification pipelines. Requires @standardbeagle/[email protected]+ (dartai_loop_snapshot tool used by /dartai:start).
29 skills · 6 agents · hooks
Token-efficient Figma integration with query-based data shaping, image export, design tokens, and adversarial design library extraction. Build production-ready component libraries with exact CSS, assets, HTML mockups, and comprehensive documentation using the Ralph Wiggum quality loop.
28 skills · 4 agents
Contextually-loaded development standards and project-aware skills. Replaces bulk CLAUDE.md instructions with on-demand rules, language-scoped guidelines, and project-type skill templates.
25 skills · hooks
SLOP (Simple Language Open Protocol) integration for MCP management. Migrate existing MCP setups, search and execute tools, manage resources, and orchestrate multiple MCP servers. Includes Dart MCP tool reference skills for reduced get_metadata calls.
23 skills · 2 agents · 1 MCP server
UX-driven development toolkit for building accessible, user-centered websites and applications. Integrates UX principles, accessibility verification (WCAG 2.2), Nielsen's heuristics, and usability best practices directly into the development workflow. Includes agents for comprehensive audits and component design.
20 skills · 4 agents
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
15 skills · hooks
Talk like caveman. Cut filler. Keep technical details explicit.
22 skills · 1 command · 3 agents · hooks
Persistent memory for Claude Code: the next session starts where the last one left off.
29 skills · 1 MCP server · hooks
Up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your coding agent's context.
1 MCP server
AI memory system — mine projects and conversations into a searchable palace. 47 MCP tools, auto-save hooks, guided setup.
3 skills · 6 commands · 1 MCP server · hooks
Build powerful, stateful CLI interfaces for any GUI application using the cli-anything harness methodology.
5 commands
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.