Thai-first context compression: shrinks large tool outputs (Thai and English) and keeps every original one sarupretrieve away. Needs uv; Ollama is optional (run /sarup:sarup-setup).
★not rated 0 20d agoA
tokens not measured
originalMIT
Publish to Instagram, TikTok, X, LinkedIn, Bluesky, Threads, Facebook and YouTube Shorts from Cursor: check a draft against every network's limits, publish or schedule it, and read the receipt that says what actually landed and what failed.
★not rated 0 19d agoA
tokens not measured
originalMIT
Record your screen and export studio-styled demo videos on macOS — wallpapers, device and photo mockups, auto-zoom, cinematic effects — driven by your AI assistant. Runs locally; requires the free Screenify Studio app.
★not rated 0 19d agoA
tokens not measured
originalMIT
USB hardware discovery for your AI assistant: find maker boards (Arduino, ESP32, Pico), their serial ports and chips, plus skills for bring-up, serial debugging, crash decoding and safe flashing.
★not rated 0 18d agoA
tokens not measured
originalMIT
Analyze JVM thread dumps without leaving Claude Code: deadlocks, pool exhaustion, CPU-bound threads and virtual thread pinning, with a health score and a shareable report. No account needed.
★not rated 0 17d agoA
tokens not measured
originalMIT
Residential and mobile proxies for AI agents: Pool Gateway proxy URLs, live stock per country, IP rotation, Private modems, usage, billing and support, through the @proxies-sx/mcp-server MCP server.
★not rated 0 13d agoA
tokens not measured
originalMIT
Deploy anything Claude makes to the web in seconds. Publish pages, sites, prototypes, reports, and PDFs to a shareable link, then update them, add a password or expiry, put them on your own domain, check visits, roll back, or get a QR code.
★not rated 0 18d agoA
tokens not measured
originalMIT
Competitive and market intelligence for B2B SaaS: competitor profiles, strategic moves, pricing changes and market landscapes, with every fact dated and sourced. Free public tier, no account needed.
★not rated 0 18d agoA
tokens not measured
originalMIT
Security copilot for AI coding agents. 10 real scanners your agent can call: SAST (441 rules, 7 languages), secrets (419 patterns), CVEs via OSV.dev, MCP/skill vetting, MITRE ATT&CK, threat modeling. Rust-native, deterministic, 100% local, no account, no API key.
★not rated 0 16d agoA
tokens not measured
originalApache-2.0
TestMu AI (formerly LambdaTest) plugins for Claude Code: Kane CLI for browser and mobile testing, Rook for testing AI agents, and the TestMu AI MCP server.
★not rated 0 3d agoA
tokens not measured
originalMIT
Universal memory runtime — ambient working-memory substrate for Claude Code. Memory is not a feature you invoke — it's the substrate the agent operates on.
Find and vet paid x402 APIs before an agent pays: an MCP server and a skill for a registry of pay-per-call services whose delivery is checked by buying them with real USDC.
★not rated 0 13d agoA
tokens not measured
originalMIT
Brazilian stock data from public CVM and B3 sources: fundamentals, end-of-day prices, dividends, total return, FIIs, ETFs and BDRs, through the remote Dados B3 MCP server.
★not rated 0 9d agoA
tokens not measured
originalMIT
Give your agent a real phone number. Send and read SMS, iMessage and WhatsApp, place AI voice calls, read call transcripts, and get a daily briefing of who reached you.