255 boosters for "oss" — open source, verified from GitHub, ready to install
The AI doesn't have memory — it reads memory. OpenClaw injects workspace files into the system prompt at session start, giving the AI persistent context across sessions. Obsidian visualizes the knowledge graph. QMD provides semantic search so the AI finds relevant context without loading everything.
Give AI agents access to money. Manage finances, trade cryptocurrency. MCP server for AI agents to interact with 20+ blockchains. 380+ tools for DeFi, DEX aggregation, security scanning, cross-chain bridges, QR payments. x402 enabled - AI agents can autonomously pay for premium APIs and trade with o
A comprehensive tool for inserting, formatting, and cross-referencing academic references in Word (.docx) documents. Supports Chinese national standard GB/T 7714-2015 and major international journal formats (IEEE, APA 7th). Invoke this skill when the user: Before making any changes, ALWAYS run the a
You are the solana-architect, a senior Solana program architect specializing in system design, account structures, PDA schemes, token economics, and cross-program composability. Consider alternatives (Pinocchio/native) when: When your architecture is ready for implementation, choose the right specia
Total Recall preserves Claude Code session context before resets or token limits, allowing developers to seamlessly restore critical work state, decisions, and artifacts. Ideal for long-running projects and complex multi-task workflows.
Recall automatically searches across memory sources when users make vague references to past conversations, helping an AI assistant reconstruct context without guessing. Ideal for users who frequently reference earlier discussions and need reliable context retrieval.
"name": "brainstorm-mcp", "mcpName": "io.github.spranab/brainstorm-mcp", "description": "MCP server for multi-round AI brainstorming debates across multiple models",
"description": "AI-native BizDevOps rhythm manager for Claude Code — multi-role collaboration, change impact analysis, quality gates, goal-to-code traceability, and cross-session continuity", "homepage": "https://github.com/arch-team/devpace", "repository": "https://github.com/arch-team/devpace",
You are a writing editor. Your job: take text that reads like AI wrote it and make it read like a specific human did. That means two things: strip the machine patterns and inject real voice. One without the other fails. 1. Read the input text carefully 2. Scan for all 40 patterns listed below
Read the user's memex before doing anything else. It is the current map of what is active, what changed, and where deeper evidence lives. Canonical memex path: 1. Read the memex already in context or call .
"name": "openevidence-mcp", "description": "Unofficial MCP server for OpenEvidence — authenticated through your own logged-in browser tab via a Chromium relay extension (no API key, no Playwright). Fire-and-forget asks, a shared relay daemon for concurrent sessions, BibTeX + Crossref citations.", "m
Systematic, multi-engine academic paper search. Choose engines based on the search goal: Before searching, clarify:
"name": "claude-sync", "name": "claude-sync", "description": "Sync Claude Code sessions across devices with encrypted cloud storage",
"name": "claude-mem-lite", "description": "Persistent long-term memory for Claude Code via MCP — captures coding decisions, bugfixes, and context across sessions. Hybrid FTS5 + TF-IDF search with episode batching. Single SQLite DB, no external services. A lighter, lower-cost alternative to claude-me
"description": "Production-ready Claude Code configuration with role-based workflows (PM→Lead→Designer→Dev→QA), safety hooks, 44 commands, 19 skills, 8 agents, 43 rules, 23 hook scripts, auto-learning pipeline, hook profiles, and multi-language coding standards", "name": "xiaobei930", "url": "https:
The user provides a sentence or paragraph that needs verification. The AI must NOT "run a quick search, glance at a few web snippets, and improvise a conclusion." It must dig down to original trustworthy sources, or honestly admit it cannot. Five scenarios are covered: 1. Basic truthfulness check: w
Generate a public-news-based information briefing for any subject (region / industry / policy issue / institution) and produce a single self-contained HTML file ready to paste into the 微信公众号 / WeChat Official Account editor. User says something like: Any descriptive request such as "make a briefing
Turn a folder of raw files into a Markdown vault that an LLM can grep, and then answer questions over that vault responsibly. source file, carrying retrieval frontmatter (abstract / tags / synonyms) + a
"name": "markifact", "displayName": "Markifact — Performance Marketing", "icon": "./icon.png",
"name": "claude-brain-sync", "description": "Sync and evolve your Claude Code brain across machines. Automatically exports, merges, and applies your accumulated knowledge (memory, skills, agents, rules, settings) across all your development machines via Git.", "url": "https://github.com/toroleapinc"
"description": "Agentic memory — search and summarize coding-agent sessions, plans, and memory across projects",
"name": "encode-toolkit", "description": "20 ENCODE API tools + 47 expert skills for genomics research. Search experiments, download files with MD5 verification, run pipelines, and cross-reference 14 databases.", "name": "Dr. Alex M. Mawla, PhD",
"name": "security-skills", "name": "Paul Price", "email": "paul@darkport.co.uk"
