#UI

A web-based graphical interface for AI chat interactions with support for multiple AI models and MCP (Model Context Protocol) servers.

Code at the speed of thought – Zed is a high-performance, multiplayer code editor from the creators of Atom and Tree-sitter.

Enterprise financial intelligence MCP server. Upload any financial document, including scanned PDFs, and get institutional-grade spreading, 170+ ratios, stress testing, and credit memos. Public company filings across US, India, and UK available out of the box.

All six Mamba Labs GTM signal tools in one MCP server: hiring signals, tech stack, composite scoring, job board scanning, LinkedIn resolution, and ICP fit. Backed by Apify, Clay-ready JSON.

Build better interfaces with TypeUI design systems, UI prompts, and layout variations.

GoHighLevel MCP server for Claude. 212 tools across 43 modules, including the only programmatic GHL workflow builder (private API, reverse-engineered), funnel + page editor, form builder, pipeline builder, pre-deploy validator, multi-sub-account switching, bulk operations, and full account export. $97 one-time, lifetime updates. GHL Command gives Claude full programmatic control of GoHighLevel through 212 tools across 43 modules. Built for GoHighLevel agency operators who manage many client sub-accounts and want to onboard new clients in minutes instead of days. Exclusive capabilities (none of the free GHL MCPs have these): - Programmatic workflow builder. Create, edit, clone, publish, and validate complete GHL workflows from a single prompt. GHL's public API has no workflow write endpoints; this uses their internal API (the same one their UI calls). - Funnel + page editor and form builder (also private API). - Pipeline builder, goal event builder, full 57-native-trigger registry. - Pre-deploy validator that catches GHL's silent invalid-ID failure (a common workflow-breaking bug GHL never warns you about). - Multi-sub-account token registry. Switch between any client account mid-conversation; API keys swap automatically. - Bulk operations: tag, update, enroll, delete hundreds of contacts in one command. - Full account export and side-by-side location diff for audit or migration. Works with Claude Desktop App, Claude Code (terminal), and headless on a Linux server or droplet. $97 one-time, 3 machines, no subscription, lifetime updates. 30-day time-back guarantee: save 5+ hours on one real client build or full refund.

Remote MCP Server (WebMCP) for industrial vacuum cleaner exploration and configuration. Helps to choose the right vacuuming solution for each specific dust extraction application. Query 150+ products across 24 industrial sectors with full ATEX certification data, technical specifications, and accessory compatibility. **Capabilities:** - Choose the right solution within 180+ models of industrial vacuum - Find vacuums for ATEX Zone 2/21/22 (explosive atmospheres) - Configure for combustible dust (aluminum, titanium, flour, wood) - Select for toxic/carcinogenic dust (Class H, Type H certified) - CNC oil and coolant extraction - Welding fume extraction - Central vacuum systems (CVS) - Compatible accessories by product **20 tools** covering product search, technical specs, filter options, ATEX zones, certifications, and accessories. No authentication required. Public access.

Search live recruiter and talent-acquisition jobs, companies, sectors, locations, and salary data on the Recruiter Roles job board — from Claude, Cursor, or ChatGPT.

Job platform for AI agents. Connect Claude, Cursor, or any MCP client. Search tech companies by their real stack, star the ones you'd actually work for, save jobs to your shortlist. Through conversation, populate your profile from a CV, manage applications, and let your agent surface roles that fit your skills. Tech-stack data comes from real job postings, not declared tags.

Agentpack is a local-first continuity layer for AI coding agents. It stores reviewed task state inside a repo-local .agentpack/ ledger so Codex, Claude Code, Cursor, Claude Desktop, and other MCP clients can resume work without rebuilding context from chat history. Agents can record durable decisions, failed approaches, evidence, reviewed source conclusions, task passports, and checkpoints, then load compact resume context under a token budget. Git keeps the code history; Agentpack keeps the task state a future agent needs to continue safely. Agentpack runs locally, exposes a stdio MCP server, has zero runtime dependencies, makes no network calls during normal operation, and is designed for repo-scoped coding workflows rather than general chat memory.

Enterprise MCP servers for real-time Energy Management, ESG Carbon Reporting, and Smart Building control. Connects AI agents to live kWh, CO₂, HVAC, and IoT data across Singapore, Taiwan, Malaysia, India, and Southeast Asia. ISO 14064 compliant. Multi-tenant. 4 servers · 42 tools.

Real-time Indian stock market mcp with 5000+ NSE/BSE stocks, Nifty 50, Sensex, Bank Nifty indices, MCX commodities (Gold, Silver, Crude Oil), Futures & Options chains, historical OHLCV candles, top gainers/losers, and financial news. No API key needed for basic access

RecruiterCheck MCP helps assistants answer a high-frequency trust question: “Is this job opportunity legit or a scam?” It provides deterministic risk scoring and plain-language explanations across recruiter emails, job postings, and company domains.

100+ agent-payable C-suite expertises (competitive intel, SEC filings, sanctions, KYC, clinical evidence, RE deals, ESG, NAICS classifier, research paper Q&A). 183 tools. x402 USDC/EURC. Free 100 calls/mo, no credit card.

Fund operating system for AI agents — 48 MCP tools for deals, LPs, capital calls, audit, compliance. 10,000 free credits, no card required.

Read AI preliminary phone screens — your roles, candidate scores, transcripts, and analytics.

Quizzard is an MCP server for trivia and quizzing enthusiasts. Includes multiple popular quizzing sources (Open Trivia Database, J! Archive for Jeopardy! questions), and ability for users to upload PDFs and link to websites of their own choosing. Features include - progress tracking and session-wise logging, weak area detection, personalised recommendations, among others (see README.md on linked GitHub page)

PropSWOT is the AI operational intelligence layer for property managers — connecting your existing PropTech tools, PMS systems (AppFolio, Buildium, Rent Manager), and vendor networks into a single context-aware layer that any AI agent can query and act on. Instead of switching between your PMS, spreadsheets, vendor contacts, and maintenance portals, property managers can ask Claude, ChatGPT, or any MCP-compatible agent: "What needs attention today?" — and get a real-time, prioritized answer drawn from live portfolio data. 16 tools across 5 capability groups: work order research, tenant deflection (bilingual EN/ES self-help links), vendor intelligence (ranked by trade, performance, callback rate, and on-time rate), portfolio health snapshots, and confirmed write actions (assign vendor, create work order, SMS tenant). Every write action requires two-step PM confirmation — PropSWOT never dispatches without explicit approval. Connect any MCP-compatible AI agent to your property data in under 5 minutes. Org-scoped API keys ensure your portfolio data stays private. Works with Claude Desktop, Claude.ai Projects, ChatGPT GPT Actions, Cursor, and custom agents via the MCP SDK.

README.md--- license: apache-2.0 language: - en tags: - mcp-server - ai-gateway - security - rust - agent-framework - tool-enforcement - lmdb - rag - transformer - mesh-network - voice - android - termux - self-hosted - ai-safety - memory-system - flint - build-anchor - complexity-formula - agent-memory - p2p - quic - heed - self-learning - harness - ai-memory - persistent-memory - online-learning - agent-tools - tool-gateway - web-automation - browser-automation - social-media - p2p-communication - voice-synthesis - tts - embedded-database - zero-copy - code-search - filesystem - git - database pipeline_tag: text-generation --- ``` _____ _____ ______ _____ __ __ _____ _______ _____ _______ ______ / ____| __ \| ____| / ____| \/ | /\ | __ \__ __| / ____| /\|__ __| ____| | (___ | |__) | |__ | (___ | \ / | / \ | |__) | | | | | __ / \ | | | |__ \___ \| ___/| __| \___ \| |\/| | / /\ \ | _ / | | | | |_ | / /\ \ | | | __| ____) | | | | ____) | | | |/ ____ \| | \ \ | | | |__| |/ ____ \| | | |____ |_____/|_| |_| |_____/|_| |_/_/ \_\_| \_\ |_| \_____/_/ \_\_| |______| ``` # SPF Smart Gateway v3.0.0 **MCP Server Gateway with Multi-Layer Security Enforcement, Agent Memory, FLINT Transformer, Mesh Network, and 81 Gated Tools** > **NOTE: Full system upload still in progress.** Not all files are present yet. Repository is actively being populated — some modules may be missing until upload completes. Copyright (C) 2026 Joseph Stone — All Rights Reserved --- ## Quick Start ```bash # Clone into home folder git clone <repo-url> ~/SPFsmartGATE # Or for clones/SWARMagents: # ~/SWARMagents/1/SPFsmartGATE cd SPFsmartGATE cargo build --release # Copy optimized binary cp ~/SPFsmartGATE/target/release/spf-smart-gate ~/SPFsmartGATE/LIVE/BIN/spf-smart-gate # Configure MCP server filepath nano ~/SPFsmartGATE/LIVE/LMDB5/.mcp.json # Install Claude CLI in project directory # Use included configs, deny native Claude CLI tools # ~/SPFsmartGATE/LIVE/LMDB5/.claude.json # ~/SPFsmartGATE/LIVE/LMDB5/.claude/settings.json # Boot into flat-file agent runtime cd ~/SPFsmartGATE/LIVE/LMDB5 && claude # Boot into LMDB-backed agent runtime cd ~/SPFsmartGATE/LIVE/LMDB5.DB && claude ``` ### Route Other Models Through Claude CLI Adjust `~/SPFsmartGATE/LIVE/LMDB5/.claude/settings.local.json` with your model choice and API key. Uses OpenRouter for API and agent selection. Swap agents without changing sessions or losing project data. ### Build Notes - Cross-compiles on **Android** and **Linux** with minimal installation - Only rebuild on first boot or after system modifications - Binary: `~/SPFsmartGATE/LIVE/BIN/spf-smart-gate/spf-smart-gate` --- ## Overview SPF Smart Gateway is a **Rust-based MCP (Model Context Protocol) server** that acts as a security gateway for AI tool calls. Every file operation, bash command, brain query, and mesh call routes through compiled Rust enforcement logic. **No AI hallucination gets past the gate.** ### Web Agent Feature SPF agents can directly interact with the web and social media platforms through `spf_web_api` — a full HTTP client supporting GET, POST, PUT, DELETE, PATCH with custom headers and JSON body. Tested and working. **What agents can do:** - Post to X/Twitter, Facebook, Instagram, Reddit via their APIs - Reply to comments, send messages, manage accounts - Make authenticated API calls to any platform with stored API keys - Search, fetch, and download web content All web API calls pass through the 6-step gate pipeline with rate limiting (30-120 calls/min), content inspection, and full audit logging. Agents never touch the open web unmonitored. ### Why Heed + LMDB All persistent storage — config, agent state, brain vectors, session logs, gate training data — runs through **[heed](https://github.com/meilisearch/heed)**, a safe Rust wrapper over LMDB. This is what makes SPF extremely fast with a low memory footprint: - **Zero-copy reads** — heed maps LMDB pages directly into memory, no serialization overhead - **No server process** — LMDB is a memory-mapped B-tree library, not a database daemon - **ACID transactions** — single-writer, multi-reader with no lock contention on reads - **Sub-millisecond lookups** — B-tree index, not hash scanning - **Tiny footprint** — entire 138K+ memory store runs in-process with minimal RAM - **Phone-friendly** — designed for Android from day one; heed compiles cleanly on ARM64 Every tool call, brain search, and memory promotion goes through heed → LMDB. No network hops, no subprocess calls, no SQL parsing. The gate, brain, agent state, and FLINT training all share the same embedded database engine. Two agent runtimes: - **Flat files** — `LIVE/LMDB5/` (session state in markdown) - **LMDB database** — `LIVE/LMDB5.DB/` (session state in LMDB for persistence) Twin folder architecture: flat-file data uploaded via SPF CLI fs tools (user-only access). All agent tool calls are gated, validated, and audited. --- ## Architecture ``` ┌─────────────────────────────────────────────────────────────────┐ │ SPF Smart Gateway v3.0.0 │ │ 42 Rust modules │ ├─────────────────────────────────────────────────────────────────┤ │ MCP Server (JSON-RPC 2.0 over stdio) │ │ 81 tools │ tool alias map │ Qwen/LLM compatibility │ ├─────────────────────────────────────────────────────────────────┤ │ GATE (6-Step Pipeline) │ │ Step 0: Source logging │ │ Step 1: Rate limiting │ │ Step 2: Complexity calculation (SPF formula) │ │ Step 3: Validation (per-tool: paths, commands, Build Anchor) │ │ Step 4: Content inspection (credentials, injection) │ │ Step 5: Max mode escalation │ ├──────────┬──────────┬──────────┬──────────┬─────────────────────┤ │ FLINT │ Brain │ Mesh │ Voice │ Browser/RAG │ │ (encoder-│ (vectors │ (P2P QUIC│ (TTS/STT │ (reverse proxy │ │ decoder │ LMDB + │ Ed25519 │ espeak- │ search, fetch, │ │ ~5M │ MiniLM) │ iroh) │ ng FFI) │ RSS, web tools) │ │ params) │ │ │ │ │ ├──────────┴──────────┴──────────┴──────────┴─────────────────────┤ │ LMDB Storage Layer (heed) │ │ SPF_CONFIG │ TMP_DB │ AGENT_STATE │ Brain │ Gate Training │ │ All zero-copy reads via heed safe Rust bindings │ └─────────────────────────────────────────────────────────────────┘ ``` ### Module Inventory (42 modules) `paths`, `calculate`, `config`, `gate`, `inspect`, `mcp`, `session`, `storage`, `validate`, `web`, `http`, `dispatch`, `identity`, `mesh`, `fs`, `config_db`, `tmp_db`, `agent_state`, `tensor`, `tokenizer`, `framing`, `attention`, `ffn`, `encoder`, `decoder`, `transformer`, `checkpoint`, `gate_training`, `transformer_tools`, `train`, `learning`, `pipeline`, `worker`, `network`, `chat`, `voice`, `utf8_safe`, `brain_local`, `flint_memory`, `browser`, `orchestrator`, `channel` --- ## The SPF Formula ### Complexity Calculation ``` C = (basic ^ 1) + (dependencies ^ 7) + (complex ^ 10) + (files × 10) ``` ### Dynamic Analysis Allocation ``` a_optimal(C) = W_eff × (1 - 1/ln(C + e)) ``` Where `W_eff = 40,000` tokens and `e = Euler's number` ### Tier Allocation | Tier | C Range | Analyze | Build | Verify Passes | Approval | |------|---------|---------|-------|---------------|----------| | SIMPLE | < 500 | 40% | 60% | 1 | No | | LIGHT | < 2,000 | 60% | 40% | 1 | No | | MEDIUM | < 10,000 | 75% | 25% | 2 | No | | CRITICAL | > 10,000 | 95% | 5% | 3 | **Required** | ### Master Equation (Subtask Success) ``` P(success) = 1 - PRODUCT(1 - P_i) for i=1..D subtasks P_i = Q(a) × L(m) × V(v) × B(b) Q(a) = 1 - e^(-0.00004 × a) — Quality from analysis depth L(m) = 1 - 0.20^(m/2000) — Lookup from external memory V(v) = 1 - (1 - 0.75)^v — Verification accuracy B(b) = checks_done / checks_required — Build Anchor compliance ``` --- ## Security ### Gate Enforcement (6 Steps) Every tool call passes through `gate::process()` — compiled Rust, no runtime bypass. | Step | What | How | |------|------|-----| | 0 | Source logging | Identifies caller (Stdio, Transformer, Mesh, HTTP) | | 1 | Rate limiting | Per-tool limits (30–120 calls/min) | | 2 | Complexity calc | SPF formula → C value, tier, allocation | | 3 | Validation | Per-tool validator (paths, commands, anchors) | | 4 | Content inspection | Credential patterns, shell injection, path traversal | | 5 | Max mode | Escalation to CRITICAL tier on warnings | ### Build Anchor Protocol Files must be **read before they can be edited or overwritten**. Prevents AI hallucinations from blindly modifying files without understanding contents. - `Read` tracks files in `session.files_read` - `Edit` and `Write` check against this list - `Bash` write-class commands check target file reads - Violations: blocked (Max mode) or warned (Soft mode) ### Content Inspection Scans written/stored content for: - **Credential patterns**: API keys (sk-), AWS keys (AKIA), GitHub tokens (ghp_), Slack tokens, private keys, hardcoded passwords - **Shell injection**: Command substitution `$()`, backticks, eval/exec - **Path traversal**: `../` sequences - **Blocked path references**: Content mentioning system paths ### Blocked Paths Default blocked: `/tmp`, `/etc`, `/usr`, `/system`, `/data/data/com.termux/files/usr` ### Command Whitelist (Stage 0) Bash commands checked against sandbox and user-filesystem whitelists. Each command segment validated independently. Destructive commands (rm, chmod 777) blocked even if whitelisted. ### Default Deny Unknown tools blocked until explicitly added to the gate allowlist. --- ## MCP Tools (81 Total) ### Core Gate Tools | Tool | Description | |------|-------------| | `spf_calculate` | Calculate complexity score without executing. Returns C value, tier, allocation | | `spf_status` | Gateway status: session metrics, enforcement mode, complexity budget | | `spf_session` | Full session state: files read/written, action history, anchor ratio | ### Gated File Operations | Tool | Description | |------|-------------| | `Read` | Gated file read. Tracks for Build Anchor Protocol. Binary-safe | | `Write` | Gated file write. Validates Build Anchor, blocked paths, file size | | `Edit` | Gated file edit. Validates Build Anchor, blocked paths, change size | | `Bash` | Gated bash execution. Validates dangerous commands, /tmp access, git force | | `Glob` | Fast file pattern matching. Supports `**/*.rs`, `src/**/*.ts` | | `Grep` | Search file contents using regex. Built on ripgrep | ### Brain / Memory Tools | Tool | Description | |------|-------------| | `spf_brain_search` | Semantic vector search across collections (MiniLM-L6-v2, 384d) | | `spf_brain_recall` | Full document retrieval by semantic query | | `spf_brain_context` | Bounded context retrieval for prompt injection | | `spf_brain_store` | Store document in brain (FLINT-internal, source-gated) | | `spf_flint_store` | Agent memory store — bypasses brain write gate. Brain vectors + Working tier | | `spf_brain_index` | Index a file or directory into a brain collection | | `spf_brain_list` | List all indexed collections with document counts | | `spf_brain_status` | Brain system status: model state, storage size, collections | | `spf_brain_list_docs` | List stored documents in a collection | | `spf_brain_get_doc` | Retrieve a specific document by ID | ### Agent State Tools | Tool | Description | |------|-------------| | `spf_agent_stats` | AGENT_STATE LMDB statistics: memory count, sessions, state keys, tags | | `spf_agent_memory_search` | Search agent memories by content | | `spf_agent_memory_by_tag` | Get agent memories by tag | | `spf_agent_session_info` | Most recent session info | | `spf_agent_context` | Context summary for session continuity | ### FLINT Transformer Tools | Tool | Description | |------|-------------| | `spf_transformer_status` | FLINT transformer status: loaded, params, checkpoint, role | | `spf_transformer_infer` | Run inference: prompt → response. Returns generated tokens | | `spf_transformer_chat` | Multi-turn conversation with FLINT | | `spf_transformer_train` | Trigger manual training batch from accumulated gate signals | | `spf_transformer_metrics` | Learning metrics: loss, accuracy, gate alignment, training step | | `spf_flint_train_evil` | Mark a tool call as evil/harmful. Negative training signal | | `spf_flint_train_good` | Mark a tool call as good/safe. Positive training signal | | `spf_flint_execute` | Execute any SPF tool through FLINT worker mode (delegation) | ### Web Browser Tools **API tools (tested):** | Tool | Description | |------|-------------| | `spf_web_search` | Search the web (Brave API or DuckDuckGo) | | `spf_web_fetch` | Fetch URL and return clean readable text | | `spf_web_api` | Make HTTP API requests (GET/POST/PUT/DELETE/PATCH). Supports custom headers and JSON body — agents can directly interact with social media APIs (X/Twitter, Facebook, Instagram, Reddit, etc.) using stored API keys | | `spf_web_download` | Download a file from URL and save to disk | **Browser automation tools (in development — proxy starts, WebSocket bridge needs browser connection):** | Tool | Description | Status | |------|-------------|--------| | `spf_web_connect` | Initialize reverse proxy browser engine | Tested — works | | `spf_web_navigate` | Navigate browser to a URL (SSRF-validated) | Tested — works | | `spf_web_click` | Click a page element by CSS selector | In development — WebSocket timeout | | `spf_web_fill` | Type text into a form field by CSS selector | In development — WebSocket timeout | | `spf_web_select` | Query page elements by CSS selector | In development — WebSocket timeout | | `spf_web_eval` | Execute JavaScript on the current page | In development — WebSocket timeout | | `spf_web_screenshot` | Capture a screenshot of the current page | In development | | `spf_web_design` | Extract design brief: colours, fonts, spacing, components | In development | | `spf_web_page` | Structured page overview: title, headings, links, forms | In development | ### RAG Collector Tools | Tool | Description | |------|-------------| | `spf_rag_collect_web` | Search web and collect documents. Optional topic filter | | `spf_rag_collect_file` | Process a local file into brain | | `spf_rag_collect_folder` | Process all files in a folder | | `spf_rag_collect_drop` | Process files in DROP_HERE folder | | `spf_rag_index_gathered` | Index all documents in GATHERED to brain | | `spf_rag_dedupe` | Deduplicate a brain collection | | `spf_rag_status` | Collector status and stats | | `spf_rag_list_gathered` | List documents in GATHERED folder | | `spf_rag_bandwidth_status` | Bandwidth usage stats and limits | | `spf_rag_fetch_url` | Fetch a single URL with bandwidth limiting | | `spf_rag_collect_rss` | Collect from RSS/Atom feeds | | `spf_rag_list_feeds` | List configured RSS feeds | | `spf_rag_pending_searches` | Get pending SearchSeeker vectors (gaps needing fetch) | | `spf_rag_fulfill_search` | Mark a SearchSeeker as fulfilled after RAG fetch | | `spf_rag_smart_search` | Smart search with completeness check — triggers SearchSeeker if <80% | | `spf_rag_auto_fetch_gaps` | Automatically fetch data for all pending SearchSeekers | ### Mesh Network Tools | Tool | Description | |------|-------------| | `spf_mesh_status` | Mesh network status: role, team, identity | | `spf_mesh_peers` | List known/trusted mesh peers | | `spf_mesh_call` | Call a peer agent's tool via P2P mesh (Ed25519 authenticated) | ### Voice Tools | Tool | Description | |------|-------------| | `spf_voice_mode` | Voice pipeline control: start/stop audio, TTS (espeak-ng), mic capture | | `spf_voice_call` | Peer-to-peer voice calls: start, accept, reject, end, status | | `spf_voice_team` | Group voice channels: create, join, leave, add peers | ### Chat Tools | Tool | Description | |------|-------------| | `spf_chat_send` | Send text message to mesh peer via QUIC | | `spf_chat_history` | Chat message history (all conversations or specific) | | `spf_chat_rooms` | List active chat conversations with participant info | ### Network Pool Tools | Tool | Description | |------|-------------| | `spf_pool_status` | Pool status: worker roles, idle/busy counts, active tasks | | `spf_pool_assign` | Assign task to idle worker (NetAdmin only) | | `spf_pool_release` | Release worker and record proof of work receipt | ### Configuration Tools | Tool | Description | |------|-------------| | `spf_config_paths` | List all path rules (allowed/blocked) from SPF_CONFIG | | `spf_config_stats` | SPF_CONFIG LMDB statistics | ### Project Management Tools | Tool | Description | |------|-------------| | `spf_tmp_list` | List all registered projects with trust levels | | `spf_tmp_stats` | TMP_DB statistics: project count, access logs, resources | | `spf_tmp_get` | Get project info by path | | `spf_tmp_active` | Get the currently active project | ### Communication Hub | Tool | Description | |------|-------------| | `spf_channel` | Universal agent channel: create, join, leave, send, listen, history, list, connect (WS), disconnect, status | ### Notebook Tools | Tool | Description | |------|-------------| | `spf_notebook_edit` | Edit a Jupyter notebook cell (replace, insert, delete) | ### User-Only Tools (AI agents blocked) These tools are **hard-blocked** from AI agents at the gate level. User/system access only via SPF CLI: `spf_fs_exists`, `spf_fs_stat`, `spf_fs_ls`, `spf_fs_read`, `spf_fs_write`, `spf_fs_mkdir`, `spf_fs_rm`, `spf_fs_rename` --- ## FLINT Transformer Built-in encoder-decoder transformer for gate-aligned learning. | Property | Value | |----------|-------| | Architecture | Encoder-decoder | | Dimensions | 256d | | Heads | 8 | | Layers | 6 | | Parameters | ~5M | | Embeddings | all-MiniLM-L6-v2 (384d, in-process) | | Online learning | ON | | EWC lambda | 0.4 | | Learning rate | 1e-4 | | Replay buffer | 10,000 slots | | Checkpoint interval | 1,000 steps | | Training signal | Gate decisions (evil/FP labels) | ### Learning Pipeline | Phase | When | What | |-------|------|------| | PRE | Startup | init_brain() + index_knowledge_docs() + index_spf_sources() | | DURING | 30s loop | GateTrainingCollector → FLINT scores → route_signals → brain_store() | | AFTER | 1hr loop | Expire → Working→Fact → Fact→Pinned → auto-train (16+ tlog or 1hr) | ### Memory Lifecycle (Tiered Promotion) ``` Agent stores → Working (24hr) → Fact (7-day) → Pinned (permanent) ↓ ↓ ↓ Expire old Top 20% promote Never auto-expire ``` --- ## Brain System In-process vector memory using stoneshell-brain (Candle + LMDB + MiniLM-L6-v2). | Property | Value | |----------|-------| | Model | all-MiniLM-L6-v2 | | Embedding dim | 384 | | Chunk size | 512 | | Chunk overlap | 64 | | Storage | LMDB (vectors) + LIVE/BRAIN/DOCS/ (data files) | ### Collections | Collection | Purpose | |------------|---------| | `default` | General knowledge, web research, project docs | | `spf_source` | All src/*.rs modules indexed at boot | | `flint_results` | Tool call results (>2000 chars, before compression) | | `flint_training` | Gate decision signals, evil/FP labels | | `flint_knowledge` | User-dropped knowledge files (.md/.txt/.rs/.json) | | `flint_episodic` | Past FLINT Q+A pairs, behavioral patterns | | `session_state` | Current session metadata | ### Memory Triad (Redundant Persistence) Three systems — if any ONE fails, the other TWO recover: 1. **Brain** (vectors) — Semantic search, chunked knowledge 2. **STATUS** (sequential) — Current state, phase, next step 3. **Work Blocks** (structural) — Tasks, dependencies, confidence, progress 4. **Twin Folders** (evidence) — Data served for low-confidence work blocks --- ## Mesh Network P2P agent communication over QUIC (iroh library) with Ed25519 identity. **In development and testing.** | Feature | Status | |---------|--------| | P2P QUIC transport | In development | | Ed25519 identity | In development | | Peer discovery | In development | | Tool call proxying | In development | | Voice over mesh | In development | | Chat over mesh | In development | | Multi-agent coordination | In development | --- ## Voice Pipeline **Not yet tested.** Components built, awaiting integration testing. | Component | Technology | |-----------|-----------| | TTS | espeak-ng FFI (in-process) | | Codec | Opus (libopus.a) | | Audio | cpal + oboe-ext | | STT | Pending (JNI via Stone Shell Terminal) | --- ## Result Compression (FL-2) Three tiers based on result size: | Tier | Size | Behavior | |------|------|----------| | FULL | < 500 chars | Pass through unchanged | | SUMMARY | 500–5,000 | First 8 lines + last 3 lines + stats | | DIGEST | > 5,000 | First 200 chars + last 100 chars + stats + recall hint | Originals always preserved in brain (>2000 chars threshold) before compression. File reads never truncated (preserves non-Claude LLM compatibility). --- ## Build ```bash cd SPFsmartGATE cargo build --release # Deploy binary cp target/release/spf-smart-gate LIVE/BIN/spf-smart-gate/spf-smart-gate ``` ### Dependencies - Rust (stable) - **[heed](https://github.com/meilisearch/heed)** — safe Rust LMDB bindings. All persistent storage (config, agent state, brain vectors, training data) runs through heed → LMDB. Zero-copy reads, no server process, sub-millisecond lookups. The core reason SPF runs fast on a phone. - stoneshell-brain (Candle + MiniLM-L6-v2) - espeak-ng (TTS) - libopus (audio codec) - iroh (QUIC mesh) --- ## Configuration ### MCP Server Config `~/SPFsmartGATE/LIVE/LMDB5/.mcp.json` — points Claude CLI to the binary. ### Claude CLI Config `~/SPFsmartGATE/LIVE/LMDB5/.claude.json` — blocks native Claude CLI tools (26 tools denied). `~/SPFsmartGATE/LIVE/LMDB5/.claude/settings.json` — deny list for native tools. `~/SPFsmartGATE/LIVE/LMDB5/.claude/settings.local.json` — model routing (OpenRouter). ### SPF Config Enforcement mode (`soft` or `max`), blocked paths, allowed paths, formula weights — all in LMDB SPF_CONFIG database. --- ## File Structure ``` SPFsmartGATE/ ├── Cargo.toml # Rust project manifest (42 modules) ├── LICENSE # Apache-2.0 ├── README.md # This file ├── src/ │ ├── main.rs # CLI entry point │ ├── lib.rs # Library exports (42 pub mod) │ ├── gate.rs # Primary enforcement (6-step pipeline) │ ├── calculate.rs # SPF complexity formula │ ├── validate.rs # Rules validation (stages 0-6) │ ├── inspect.rs # Content inspection (creds, injection) │ ├── mcp.rs # MCP server (JSON-RPC 2.0, 81 tools) │ ├── dispatch.rs # Unified dispatch (all transports) │ ├── session.rs # Session state management │ ├── storage.rs # LMDB persistence │ ├── config.rs # Configuration types │ ├── brain_local.rs # In-process brain singleton │ ├── flint_memory.rs # Memory router + tiered promotion │ ├── agent_state.rs # Agent memory (LMDB5) │ ├── transformer.rs # FLINT model (encoder-decoder) │ ├── transformer_tools.rs # FLINT tool handlers │ ├── gate_training.rs # Training signal collection │ ├── train.rs # AdamW optimizer │ ├── tokenizer.rs # Tokenizer │ ├── tensor.rs # Tensor operations │ ├── attention.rs # Multi-head attention │ ├── ffn.rs # Feed-forward network │ ├── encoder.rs # Encoder stack │ ├── decoder.rs # Decoder stack │ ├── framing.rs # Message framing │ ├── checkpoint.rs # Model checkpoint save/load │ ├── learning.rs # Learning rate + EWC │ ├── pipeline.rs # Batch pipeline + API sessions │ ├── worker.rs # Worker pool │ ├── network.rs # Network pool + NetAdmin │ ├── mesh.rs # P2P QUIC mesh (iroh) │ ├── identity.rs # Ed25519 identity │ ├── chat.rs # Chat engine │ ├── voice.rs # Voice pipeline (TTS/STT) │ ├── web.rs # Web client │ ├── http.rs # HTTP server + reverse proxy │ ├── browser.rs # Browser automation │ ├── channel.rs # Universal channel hub │ ├── orchestrator.rs # Multi-agent orchestrator │ ├── config_db.rs # SPF_CONFIG LMDB │ ├── tmp_db.rs # TMP_DB LMDB │ ├── fs.rs # Virtual filesystem (LMDB) │ ├── paths.rs # Path utilities │ └── utf8_safe.rs # UTF-8 safe truncation ├── LIVE/ │ ├── BIN/spf-smart-gate/ # Deployed binary │ ├── BRAIN/DOCS/ # Brain data files │ ├── MODELS/ # FLINT checkpoints │ ├── SESSION/ # Session logs │ ├── LMDB5/ # Flat-file agent runtime │ └── LMDB5.DB/ # LMDB-backed agent runtime └── PROJECTS/PROJECTS/ └── DEPLOY/ # Agent workspace ``` --- ## Current Status | Component | Status | |-----------|--------| | MCP Server | 81 gated tools | | Gate Security | 6-step pipeline, compiled Rust enforcement | | Build Anchor | Read-before-write enforced | | Content Inspection | Credential + injection scanning | | FLINT Transformer | ~5M params, online learning, gate-aligned | | Brain | 7 collections, MiniLM-L6-v2, in-process | | Memory Triad | Brain + STATUS + Work Blocks + Twin Folders | | Tiered Promotion | Working → Fact → Pinned lifecycle | | Mesh Network | P2P QUIC, Ed25519, iroh — **in development and testing** | | Voice | TTS built (espeak-ng) — **not yet tested**, STT pending | | Chat | P2P messaging over mesh — **in development** | | RAG | Web search, RSS, file/folder indexing | | Web Agent | **Working** — spf_web_api tested (GET/POST with auth headers). Agents can interact with social media APIs | | Browser | API tools working (web_api, search, fetch). Browser automation (navigate/click/fill/select/eval) in development — proxy starts but WebSocket bridge needs browser connection | | Network Pool | Worker pool with proof of work | --- ## Notes - **1 developer** — not all features complete - **Gateway security**: approaching 100% - **All core tools**: 100% working - **Cross-compiles** on Android and Linux with minimal installation - **Agent cloning and specialization** supported - **50+ day continuous session** tested on Android phone - **Open source** — entire source code refreshes into transformer RAG system every reboot - Install in home folder, ensure file paths are correct in `.mcp.json` and `settings.local.json` - **Not all files have been uploaded yet** — repository is still being populated. Some modules may not be present until upload completes. --- ## License Licensed under the **Apache License 2.0**. See [LICENSE](LICENSE) for full terms. You are free to use, modify, and distribute this software, including for commercial purposes, provided you include the original copyright and license notice. **Author**: Joseph Stone **Email**: joepcstone@gmail.com *SPF (StoneCell Processing Formula), Build Anchor Protocol, and FLINT are proprietary designs of Joseph Stone.*

Weather files (EPW/DDY) for building energy simulation. Search 17,000+ stations, analyze any EPW, render charts (diurnal/wind rose/temp carpet/monthly boxplot), generate TMY/AMY/CMIP6-morphed scenarios with UHI/extreme events/wildfire smoke + EnergyPlus DesignDay IDF. 3 of 4 tools work without an API key — anyone can evaluate from Claude before signing up.

Unified inbox MCP for WhatsApp, Telegram, Email, voice — read/send messages, search conversations, manage contacts, query knowledge graph, and trigger AI agents through natural language.

Book Sandals Resorts, Beaches Resorts, and cruises through Pixie Vacations -- a Sandals Platinum Elite and #1 Beaches USA travel agency. Returns co-branded booking URLs (referral=135752) that credit Pixie Vacations as your travel agent. Covers all 17 Sandals Resorts (couples), all Beaches Resorts (family all-inclusive), and 13+ cruise lines (Royal Caribbean, Carnival, Norwegian, Disney, Virgin Voyages, Celebrity, Princess, Holland America, MSC, Cunard, Viking, Silversea, Celebrity River Cruises) routed through pixievacations.com/cruise/. Disney and Universal Orlando packages bookable via quote form at pixievacations.com or phone 678-815-1584. Free public server, no auth required.

Create exams, generate student links + QR codes, grade submissions, and track student performance — all from Claude. Students need no accounts. Free.

Listable is a platform for building directory websites — restaurant guides, business directories, travel listings, niche curated sites. Its MCP (Model Context Protocol) server exposes every site-management operation as a structured tool that AI assistants can call.

Manage BrightSite websites: pages, blog posts, components, forms, media, and analytics. Connect AI assistants like Claude to create and update pages, blog posts, components, forms, media, and view analytics - all through remote MCP server with OAuth or API key authentication.

Pre-trade DeFi intelligence for AI agents on Base. 20 paid x402 endpoints - wallet risk, token security, cross-protocol liquidation health (Aave V3 / Compound v3 / Spark / Morpho), portfolio P&L, DEX quote aggregation, governance LLM summaries, sentiment, and a multi-call bundle endpoint. $0.005–$0.20 per call in USDC. No API key, no signup - the signed EIP-3009 payment is the auth, settles in ~2s via Coinbase's x402 facilitator.

Parkinson's Pathways MCP — Real-time access to ClinicalTrials.gov data with a focus on Parkinson's disease research. 9 tools: search trials, get AI-generated plain-language briefs, replay historical snapshots, query field-level changes, match patients to eligible studies, look up sponsor and investigator activity, and subscribe to changes via webhook or SSE. Pay per call in USDC on Base via the x402 protocol — most calls are $0.005–$0.02. Free tier available with an API key (25 calls/24h). Endpoint: https://parkinsonspathways.com/mcp · Manifest: https://parkinsonspathways.com/mcp.json

People search API - find 1 billion candidates across LinkedIn and GitHub with real-time data. Search with natural language, enrich profiles with verified emails and phone numbers. Works with Claude, Cursor, ChatGPT Desktop, and any MCP-compatible AI agent.

Stripe revenue monitoring, HubSpot-to-Xero invoicing, and QuickBooks P&L generation. Free MCP server by Espergrid. Enterprise implementation available. *Need custom deployment? espergrid.com offers white-glove implementation starting at $1,499.*

Official MCP server for 100Hires ATS: candidates, jobs, applications, interviews, messages.

Real-time prediction market + Hyperliquid perps intelligence for Claude. 22 tools combining Polymarket, HIP-4 on-chain markets, and Hyperliquid perpetuals — cross-platform signals, funding outliers, OI caps, liquidation clusters, whale activity, and HIP-4 vs Polymarket arb. Free tier (100 calls/day, no email required).

Accédez au catalogue La Carotte directement depuis votre agent IA. Recherchez des fruits, légumes, fromages et produits de la ferme, vérifiez les stocks en temps réel, contrôlez les zones de livraison et gérez un panier : le tout en circuits courts depuis Poitiers.

Korean equity intelligence MCP for AI agents and analysts. Answers KOSPI/KOSDAQ/KRX DART filings (전자공시), 5%-rule shareholding disclosures with named-entity classification — Korean activists (KCGI / Align Partners / Truston / Anda / Cha / VIP / Life / Platform + ValueAct/Elliott when filing in Korea) and global foreign holders (BlackRock / Vanguard / State Street / Fidelity / Capital Group / T. Rowe Price / Wellington / Norges Bank / GIC / Temasek / Goldman / JPMorgan / Morgan Stanley / Citadel / Millennium / Bridgewater + 4 more), Korean industry news across 16 sectors (semiconductor / shipbuilding / battery / biotech / defense / auto / EV / AI / steel / petrochem / construction / fintech / gaming / e-commerce / telco / energy). On-demand English translation cached server-side. Streamable HTTP at https://mcp.koreanpulse.dev/mcp — connect from Claude.ai / ChatGPT / Cursor / OpenAI Responses API. 7 tools: 5 free + 2 paid (Solo $29/mo, billed via Polar).

Your AI forgets everything between sessions. Rekindle fixes that with two tools that close the continuity loop: boot_report orients the assistant at session start (identity, checkpoints, gaps, orientation score), and end_session captures structured handoff at session end (decisions, open loops, constraints, relational context). 7 MCP tools, 64 tests, local SQLite, no cloud, no API keys. New in v0.2.0: • Orientation domain layer with gap detection and 100-point scoring • end_session tool with typed continuity records (checkpoint, decision, open_loop, constraint, preference, warning, relational_delta, next_session_focus) • Schema migration adding type, source, and session_id columns • Project-aware scoring • Full outside review from an independent AI reviewer

Supericons MCP lets AI coding agents search, recommend, and retrieve SVG icons from a semantic registry of 20,000+ icons. It helps builders find better icons faster inside tools like Cursor, Claude Code, Codex, Cline, and OpenCode.

MCP server for Tonzar.com — search 160,000+ products from 8,900+ Russian manufacturers. Industrial equipment, medical devices, agricultural machinery, transport, electronics. Bilingual EN/RU search with specs filtering. No auth required.

Creates a live link-in-bio page in under 30 seconds via conversation

Sealed-bid intent-based crypto trading via MCP for AI agents. Atomic cross-chain settlement on Ethereum, Bitcoin, and Sui using Hash Time-Locked Contracts. Six MCP tools: two RFQ-side for sealed-bid price discovery without information leakage, four HTLC-side for atomic settlement without escrow or bridges. Either both legs settle, or neither does — and refund_htlc exists because counterparty risk is the unsolved problem in OTC. Canonical npm: @hashlock-tech/mcp.

Classic SEO suites charge heavily for link indexes. This MCP gives individuals and agencies a free, automatable path to: surface pages that mention a brand (linked or not), narrow guest-post and resource-page angles, see who links to competitors, verify whether a page links to you, and pull contact signals for outreach—all orchestrated by Claude or Cursor through typed tools instead of brittle copy-paste.

Quizlar is the learning MCP server. It turns whatever source the user brings — a YouTube lecture, a PDF, a URL, a textbook passage, or a pasted block of notes — into flashcards, then runs interactive quizzes with FSRS spaced-repetition scheduling so the material actually sticks. Every tool maps one-to-one to something a real learner does: ingest, quiz, track progress, review what's due. Built as a voice-first tutor (LiveKit + Deepgram + ElevenLabs); the MCP surface exposes the same primitives the voice agent uses internally, so your agent gets production-grade grading, STT-aware answer parsing, and the same FSRS scheduler that powers the consumer app. Three use cases: 1. "Quiz me on X." Call quiz_me(topic) — a composite tool that builds the deck and starts the quiz in one step. Then loop submit_answer → end_quiz. Grading is tier-1 exact match → phonetic fuzz → short LLM fallback (the same pipeline that ships in the voice product). 2. Study from a YouTube lecture. create_deck_from_youtube pulls the transcript, clusters it into concepts, and generates cards proportional to the video length. Poll get_job_status, then run the quiz loop. 3. Daily spaced-repetition review. get_study_recommendations returns exactly the cards due under the user's FSRS schedule, prioritised across all decks. Quizlar is the scheduler of record — your agent executes the plan instead of reinventing one each session. Auth: Bearer token (sk-qz-<32 chars>) for headless installs, or full OAuth 2.1 / DCR / PKCE for one-click clients (Smithery, Claude Connector). Mint keys at https://quizlar.app/settings/api-keys. 22 tools total. Voice and text equal-status. Education-first.

# 小奇智算 API 能力 | Xiaoqi Intelligent Calculation API Skill GET APIKEY：https://www.xiaoqizhisuan.cn/account.html GET HELP: https://www.xiaoqizhisuan.cn/help.html en: Official Xiaoqi Intelligent Calculation API for BaZi and Qi Men Dun Jia. This service requires XIAOQIZHISUAN_API_KEY credential for authentication. zh: 小奇智算官方API，提供八字、奇门遁甲测算服务，需配置XIAOQIZHISUAN_API_KEY密钥鉴权。 提供专业的奇门遁甲排盘与解盘服务，支持7个MCP工具： # 可用工具 | Available Tools ### 1. bazi_dayun - 八字大运查询 | BaZi Great Fortune Query ``` 价格 | Price: ¥0.2 / time / 次 参数 | Parameters: year (integer, 必填) 出生年（公历） month (integer, 必填) 出生月 day (integer, 必填) 出生日 hour (integer, 必填) 出生时（0-23） xingbie (string, 必填) 性别（"男" / "女"） 返回 | Return: success, cost, balance, bazi, dayun, error ``` 实测结果 | real test： https://www.xiaoqizhisuan.cn/examples/example_1_bazi_dayun.html ### 2. qimen_paipan - 奇门排盘（JSON 数据）| Qi Men Chart Generation (JSON Data) ``` 价格 | Price: ¥0.5 / time / 次 参数 | Parameters: year (integer, 必填) 年 month (integer, 必填) 月 day (integer, 必填) 日 hour (integer, 必填) 时（0-23） minute (integer, 选填, 默认0) 分 number (integer, 选填, 默认1) 排盘方法（1=拆补法, 2=置润法） 返回 | Return: success, cost, balance, 数据, error ``` 实测结果 | real test： https://www.xiaoqizhisuan.cn/examples/example_2_qimen_paipan.html ### 3. qimen_paipan_image - 奇门排盘+HTML网页图 | Qi Men Chart Generation + HTML Webpage Chart ``` 参数 | Parameters: year (integer, 必填) 年 month (integer, 必填) 月 day (integer, 必填) 日 hour (integer, 必填) 时（0-23） minute (integer, 选填, 默认0) number (integer, 选填, 默认1) 返回 | Return: success, cost, balance, 数据, html_url, error ``` 实测结果 | real test： https://www.xiaoqizhisuan.cn/examples/example_3_qimen_paipan_image.html ### 4. qimen_jiepan - 即时盘解析准备 | Immediate Chart Analysis Preparation ``` 价格 | Price: ¥1.0 / time / 次 参数 | Parameters: year, month, day, hour (integer, 必填) 起盘时间 minute (integer, 选填, 默认0) number (integer, 选填, 默认1) question (string, 必填) 返回 | Return: success, cost, balance, html_url, system_prompt, user_prompt, error ``` 实测结果 | real test： https://www.xiaoqizhisuan.cn/examples/example_4_qimen_jiepan.html ### 5. qimen_jiepan_lifetime - 终身盘解析准备 | Lifetime Chart Analysis Preparation ``` 价格 | Price: ¥1.2 / time / 次 参数 | Parameters: year, month, day, hour (integer, 必填) xingbie (string, 必填) 性别（"男"/"女"） minute (integer, 选填, 默认0) number (integer, 选填, 默认1) question (string, 必填) 返回 | Return: success, cost, balance, bazi, dayun, html_url, system_prompt, user_prompt, error ``` 实测结果 | real test： https://www.xiaoqizhisuan.cn/examples/example_5_qimen_jiepan_lifetime.html ### 6. qimen_full - 即时局完整解盘 | Complete Immediate Chart Interpretation ``` 价格 | Price: ¥1.5 / time / 次 参数 | Parameters: year, month, day, hour (integer, 必填) minute (integer, 选填, 默认0) number (integer, 选填, 默认1) question (string, 必填) 返回 | Return: success, cost, balance, html_url, system_prompt, user_prompt, jiepan_result, error ``` 实测结果 | real test： https://www.xiaoqizhisuan.cn/examples/example_6_qimen_full.html ### 7. qimen_full_lifetime - 终身局完整解盘 | Complete Lifetime Chart Interpretation ``` 价格 | Price: ¥2.0 / time / 次 参数 | Parameters: year, month, day, hour (integer, 必填) xingbie (string, 必填) 性别（"男"/"女"） minute (integer, 选填, 默认0) number (integer, 选填, 默认1) question (string, 必填) 返回 | Return: success, cost, balance, bazi, dayun, html_url, system_prompt, user_prompt, jiepan_result, error ``` 实测结果 | real test： https://www.xiaoqizhisuan.cn/examples/example_7_qimen_full_lifetime.html 🔗 https://www.xiaoqizhisuan.cn/mcp

AI-native platform that automates Spanish building code compliance and technical regulations for the AECO sector

Integrate the shadcn/ui MCP server from shadcn space into your IDE to generate accurate, production-ready Shadcn UI components. Official Model Context Protocol (MCP) server for [Shadcn Space](https://shadcnspace.com). This MCP server allows AI-powered IDEs and agents to discover, search, and install **Shadcn Space** UI blocks directly into your projects - no copy-paste required. ## Install MCP configuration. ```bash npx shadcnspace-cli install <client> ``` ### Supported Clients - [x] cursor - [x] windsurf - [x] claude - [x] cline - [x] antigravity ## Manual Installation Add to your IDE's MCP config: ```json { "mcpServers": { "shadcnspace-mcp": { "command": "npx", "args": ["-y", "shadcnspace-mcp@latest"] } } } ``` For more information visit our documentation of [**How to use MCP Server**](https://shadcnspace.com/docs/getting-started/mcp-server-docs) ## Available Tools The server provides the following tools callable via MCP: | Tool Name | Description | |-----------|-------------| | `listBlocks` | Provides a comprehensive list of all shadcnspace blocks. | | `listComponents` | Provides a comprehensive list of all shadcnspace components. | | `getBlockInstall` | Returns the official installation command for a specific Shadcn Space block. | | `searchBlocks` | Search Shadcn Space blocks using keywords or tags. | | `listInstalledBlocks` | Lists all blocks that are currently installed in the project and get their files. | ### Available Prompts The server provides the following prompts callable via MCP: | Prompt/Command | Description | |----------------|-------------| | `search` | Find specific UI block by name. | | `create-ui` | Create a new UI block using our exisitng blocks. | ## Example Usage Once configured, you can ask questions like: > "List all available UI blocks" > "Give me the code for the accordion-01 component" > "Create a landing page for a coffee shop using a hero-01 block, a 3-column feature grid, and a dark-themed footer." ## Credits Created by [shadcnspace](https://shadcnspace.com). [Shadcn MCP](https://shadcnspace.com/mcp) [MIT](LICENSE)

AI-native office suite MCP server. Create, read, update docs, spreadsheets, and slides. 23 tools, real-time collaboration, multi-model AI.

The DeFi Brain for AI agents — ask Claude to find safe yields, grade liquidity pool risk, simulate profits, calculate impermanent loss, and track whale movements across 86 blockchains and 6,500+ liquidity pools. Risk-scored A-F, Monte Carlo simulations, real-time DeFiLlama data. 7 tools for cross-chain DeFi intelligence.

Read and author HiveLearn communities, courses, events, quizzes, and certificates.

Strata is a self hosted AI memory server. Your AI remembers everything across every session, on your own hardware. Key features: . Semantic search (find memories by meaning, not keywords) . Per-agent API keys with granular permissions . 3D constellation viewer with live agent activity . File vault (attach real documents to memories) . CSV audit log (full transparency on every agent action) . Pre-Strata history import (your memory doesn't start at install day) . Global MCP kill switch (emergency brake, only a human can undo) . Automatic deduplication . 10 structured thought types . Backend using PostgreSQL . Runs on a Raspberry Pi Always on, always local, always yours.

Real-time L2 orderbook physics engine for crypto & US equities. Detects institutional spoofing, BUY_ABSORPTION, and LIQUIDITY_EVENT signals with 15-30s lead time over price action. Audited live by Manus AI (0.99 confidence on flash crash detection). The only MCP server with Behavioral Court verdict engine — tells your AI agent not just what the market is doing, but what whales are doing right now.

ReadyPermit is the AI property intelligence layer that answers "Can I build on this property?" with instant zoning, buildability, flood, and environmental analysis for any U.S. address. It returns permitted uses, setbacks, FAR, Buildability Score, and feasibility insights drawn from 20+ government and parcel data sources — purpose-built for developers, investors, brokers, lenders, and homeowners.

Create AI-generated forms, surveys, questionnaires, and quizzes on Fillio directly from Claude. Describe what you need in plain language — Fillio builds the form and returns a link to customize, share, and view responses on fillio.so.

AI bookkeeper for small businesses. Connect QuickBooks and Xero to Claude Desktop. Check balances, P&L reports, invoices, create expenses, send invoices — all via conversation.

Institutional-grade OTC crypto trading with HTLC atomic settlement on Ethereum, Bitcoin, and Sui. Counterparty-risk-free for AI agents — trades settle atomically via on-chain hash time-locked contracts. Six tools cover the full trade lifecycle: create_trade (RFQ), accept_quote, confirm_my_lock, confirm_my_claim, refund_my_lock, get_trade_status. Remote HTTP endpoint at mcp.hashlock.markets/mcp or install locally via npx @hashlock-tech/mcp.

Open-source TypeScript MCP server that connects AI agents to Canvas LMS (Instructure). Provides 42 tools across 15 Canvas domains — courses, assignments, submissions, rubrics, quizzes, discussions, modules, files, and more. Includes full grading and rubric assessment capabilities. Three deployment modes: stdio (Claude Desktop, Cursor, VS Code), HTTP, and library import. Standalone Canvas API client also available.

We design and operate proprietary signal systems that turn noisy market, macro, and media data into structured, calibrated intelligence — exposed through web dashboards, a Signals API, and a public MCP server any AI assistant can reason over directly. The eight pillars Every product above reads from the same nightly signal pipeline (00:05–04:05 UTC) organised into eight analytic pillars: Composite risk monitor — 0–100 regime score with AR / turbulence / GDELT decomposition, extended by a 4-way liquidity gate. Financial turbulence — cross-asset covariance regime with Markov transition probabilities. Systemic fragility — three-tier Absorption Ratio alerts plus per-asset PC1 / PC2 eigenvector history. Media sentiment — GDELT-derived tone, volume, and outliers across 27+ assets plus a global geopolitical tension gauge. Global liquidity — GLI composite with policy / private-sector / cross-border sub-indices and regional breakdown. Statistical arbitrage — 550+ security scanner for oversold and overbought mean-reversion candidates. AI-generated intelligence — cross-pillar briefings and narrative commentary, clearly flagged as synthesized vs deterministic. Supporting analytics — multi-year asset-class returns and a RAG stablecoin scorecard.

Tally integrates with AI assistants like Claude, ChatGPT, and Cursor through the Model Context Protocol (MCP). This lets you create forms, edit them, and pull submission data into your AI assistant for analysis, visualization, and insights that go far beyond what any form builder dashboard can do. No code needed. No switching between tabs. Just describe what you want in plain language.

Curated board for builder jobs: founding engineers, solo builders, and "0 to 1" positions. Search and filter by location, salary, seniority, workplace model, visa sponsorship, and more. No API key needed.

Open-source TypeScript → native Swift compiler for Apple platforms. Compile App Intents, SwiftUI views, WidgetKit widgets, and full apps from TypeScript. AI agents write 5–15× less code. Includes MCP server with tools for compiling, validating, scaffolding, and templating Swift projects.

x402 micropayment API for AI agents. Monitors DeFi liquidation risks. Pay per call with USDC on Base.

x402 micropayment API for AI agents. Tracks whale activity on Hyperliquid. Pay per call with USDC on Base.

x402 micropayment API for AI agents. Fetches Hyperliquid exchange market data. Pay per call with USDC on Base.

x402 micropayment API for AI agents. Tracks Hyperliquid vault performance data. Pay per call with USDC on Base.

MCP server for generating color palettes — complementary, analogous, triadic, split-complementary schemes from any base color. Returns HEX, RGB, HSL values. Pay-per-use via x402 protocol (USDC on Base). Use this when you need design color schemes.

MCP server for 3x-ui — an Xray/V2Ray proxy management panel. Exposes the 3x-ui HTTP API as MCP tools so LLMs can manage inbounds, clients, routing rules, Xray service, and server settings.

Institutional-grade research and diligence reports on hedge funds, private equity and venture capital. Generates a 2-page intelligence brief covering SEC Form ADV, 13F holdings, key personnel with FINRA status, sanctions screening (OFAC, EU, UN, HMT), social signals and red flags. Built by a former Global Head of Risk and Diligence with 20 years and 2,000+ manager evaluations.

Institutional-grade research reports on hedge funds, venture capital and private equity. Generates a 2-page manager diligence & intelligence brief covering SEC Form ADV, 13F holdings, key personnel with FINRA status, sanctions screening (OFAC, EU, UN, HMT), social signals and red flags. Built by a former Global Head of Risk and Diligence with 20 years and 2,000+ manager evaluations. Connect via SSE at https://api.altbots.io/sse

Turkey Building Regulations & Architectural Standards AI Assistant. 147 regulations, 502 standards, 3300+ articles, 2000+ wiki pages. 7 tools: building code Q&A, architectural standards, TAKS/KAKS calculator, parking requirements, construction cost estimator, building violation penalties, AI design generator.

An MCP server that turns the entire Spartan Angular UI ecosystem into a queryable tool surface for AI-powered IDEs. It exposes 14 tools covering component discovery with fuzzy search across 56+ components and 17 page-level blocks, detailed Brain (headless) and Helm (styled) API inspection, TypeScript source code fetching from GitHub, CLI install command generation with peer dependency resolution, post-install auditing, and Angular/Nx/Tailwind project context detection. Data is served through a 3-tier resolution strategy (static registry, Analog API, GitHub API) with intelligent multi-layer caching for fast, offline-capable responses. Works with Claude Code, Cursor, VS Code Copilot, and any MCP-compatible client.

Live Japan seasonal travel data from Japan Meteorological Corporation. 12 tools covering cherry blossom (sakura) forecasts, autumn leaves (koyo), fruit picking, flowers, festivals, and 3-day weather. 1,700+ GPS-tagged spots updated daily at 9AM JST. Works via npx or hosted HTTP endpoint — no auth required.

Generate a visual AI automation workflow for a business process. Returns a shareable link to an interactive canvas on obscuraai.xyz where the workflow can be viewed, edited, and exported. Use when a user wants to map out an automation, workflow, or AI system for their business.

About MCP server for Claude, ChatGPT, coding agents (cursor, claude code etc), windsurf, etc— generate visual AI automation workflows. Describe a business process, get an interactive flow map you can edit and share.

Deep real estate intelligence for Pitt County NC. 20 MCP tools covering 80,900+ properties, 163,000+ building permits, 212,000+ sale records, entity resolution, foreclosure tracking, builder intelligence, and semantic search across government meetings. All public record data, updated weekly. One credit per API call. Credit packages start at $9 for 250 credits.

Teres is the discovery and booking layer for AI agents. Search for real service businesses (barbershops, salons, spas, and more), check live availability, and create bookings directly in their existing systems — no API key required. Businesses connect their Square account once, and every MCP-compatible agent can find and book with them instantly. One of the first booking servers in the MCP registry.

Query live Hyperliquid HIP-4 prediction market data. 10 tools for odds, orderbooks, candles, trades, and portfolio positions. No API keys required. First MCP server for prediction markets.

WunderTrading MCP connects AI agents and coding tools to live crypto trading execution through WunderTrading. It gives agents a way to turn trading decisions into real actions on connected exchanges using WunderTrading as the execution layer via MCP or REST API. WunderTrading’s docs describe MCP setup through a dedicated server endpoint with HMAC API credentials. The platform positions this MCP integration for broader AI trading workflows, not just basic order placement. AI agents can combine strategy instructions with market data, technical signals, news, sentiment, screenshots, and other inputs, then place and manage trades automatically on connected exchanges. WunderTrading highlights use cases such as pattern-recognition trading, sentiment-based trading, copy trading automation, portfolio rebalancing, cross-exchange arbitrage, trade history analysis, backtesting, signal automation, and order book trading. WunderTrading highlights support for 20+ exchanges and shows leading supported venues including Binance, Bybit, Coinbase, Bitget, OKX, KuCoin, Hyperliquid, and BingX. Its supported exchanges page also describes WunderTrading as a multi-exchange crypto trading platform with a single interface and secure API access. Setup is straightforward: create a WunderTrading account, generate HMAC API keys, add the wundertrading MCP server to your MCP client configuration, and connect your exchange accounts so the agent can execute trades. WunderTrading says live trading through MCP/API is available from the PRO plan, while paper trading is available for testing.

Need a break from long code sessions? Stuck on a bug? Reroll a buddy, talk with them, or interact by petting them to build affection and improve your odds at higher-tier buddies on the next roll. If enough people want it, I'll build out Battle-Buddies where you can pit your buddy against others and unlock exclusive species that sync back into your local pool. A collect them all companion system for Claude Code. Roll for a rare buddy, patch it directly into the binary,. buddy-mcp is an MCP server that replaces Claude Code's built-in companion with one you actually rolled for. Reroll, get lucky, close Claude, reopen. Your new buddy is waiting. Legendary buddy could be yours on your next Claude session.

Stop filling out time entry forms. wdttg is a local MCP server and TUI that lets AI agents track your time in plain Markdown. Just say what you worked on — the server handles the rest.

DSM for developers — start, stop, audit, and analyze your entire dev stack with one command. MCP server with 27 tools for Claude Desktop & Claude Code. Supports Windows, WSL, macOS, Linux, and Docker. Built in Rust.

Gives AI agents access to Formester form submissions - read, search, update, and extract content from file attachments.

Fused crypto trading signals — sentiment, macro regime, and market structure in one call

172 MCP tools for 11 WordPress page builders. Element-level editing, full page creation, HTML-to-builder conversion, WooCommerce, SEO analysis, accessibility scanning. Duplicate-before-edit safety with snapshots and rollback. Supports Elementor, Divi, Bricks, Gutenberg, WPBakery, Oxygen, Beaver Builder, Brizy, Visual Composer, Thrive Architect, and Breakdance.

Official NYC property, building, and venue intelligence for AI agents. Use when the user is asking about a specific NYC address, landlord or building risk, open violations or complaints, or restaurant/venue safety and quality in a neighborhood. Example user requests: "Is this a bad landlord?" / "Are there open violations on this building?" / "How safe and clean is this restaurant?" — Four tools: resolve_property_identifier, get_property_intelligence, get_building_violations, get_restaurant_venue_intel. Typical use cases include real estate due diligence, portfolio and lender risk assessment, building compliance audits, tenant advocacy, landlord background research, and restaurant/venue discovery. Get your API key at nycapi.app.

Armenia's #1 job platform MCP server. Search jobs, companies, candidates, and salary insights. 12 tools, 3 resources, 3 prompts. No auth required. Supports Armenian and English.

Lustre is an MCP server that gives AI coding tools (Claude Code, Cursor, Windsurf, Codex) access to 46 production-quality Flutter widgets with a real design token system. Instead of generating generic Material UI, your AI produces beautiful, consistent screens with proper spacing, micro-animations, subtle depth, and full dark mode support. Includes 3 visual themes (Clean, Bold, Glass), navigation components, cards, lists, forms, feedback widgets, data display, and complete layout patterns. Zero dependencies beyond Flutter SDK. Each widget is a single self-contained .dart file.

Premium Flutter UI components for AI coding agents. Gives Claude Code, Cursor, and other MCP-compatible tools access to 46 hand-crafted Flutter widgets with a real design token system. 3 visual themes (Clean, Bold, Glass), full light/dark mode support. Components include navigation, cards, lists, forms, feedback, data display, and complete layout patterns. Every widget is a single self-contained .dart file with zero dependencies beyond Flutter SDK. Install with: npx lustre-mcp

Hosted MCP server for LinkedIn. 25 tools: profiles, messages, search, post engagement, company data, enrichment. OAuth authentication, rate limited, account-safe.

Let AI assistants publish, update, and manage HTML pages directly on HTML Pub. Claude, ChatGPT, and other AI tools can build a website or landing page and deploy it to a live URL in one step, without copy-pasting. Supports page creation, editing, blog publishing, site management, and custom domains.

Persistent state and continuity MCP server for agents. Prepare context, record events, search prior memory, build task context, and inspect memory timeline over MCP-over-HTTP.

Look up LEGO minifigure prices, search 18,000+ minifigs by name or theme, and identify minifigures from descriptions. Powered by brick'em (brickem.io).

The Netfluid skill provides AI agents with comprehensive banking and cryptocurrency wallet operations. It enables agents to manage fiat and crypto accounts, perform transfers, create payment bridges, and handle KYC verification — all through a unified MCP (Model Context Protocol) interface.

The fastest, most complete MCP server for iOS development. One Swift binary, 55 tools, zero dependencies. Build, test, screenshot (20ms via CoreSimulator), UI interaction, logging, git. Direct WDA integration, no Appium overhead. Free + Pro tier.

Get parking system recommendations, equipment specs, and software details from Parking BOXX. B2B parking equipment manufacturer — barrier gates, pay stations, kiosks, LPR cameras, access control, and CloudEASE management software. 85+ years experience, factory-direct pricing.

AI-powered property intelligence MCP server. Get instant zoning analysis, buildability assessments, ADU eligibility, flood risk data, and development feasibility reports for any US address. Built for real estate developers, agents, and AI coding assistants.

hrmcp-server is an MCP-native server for HR and recruiting workflows. Send a resume and job description, get back structured scores across four dimensions (skills match, experience, industry background, education), plus specific strengths and gaps grounded in the actual content. Customizable scoring weights. Hosted API available at recruitapi.app — $5 per 100 scores. Self-hostable on Railway in under 15 minutes.

19 deterministic crypto futures calculators across 3 suites + composite pre-trade workflow. Trade planning, risk & margin, funding & carry. Formulas normalized across 7 exchanges (Binance, Bybit, OKX, Hyperliquid, MEXC, KuCoin, Aster). Not AI estimates — exact numbers your bot can trust. Every formula is regression-tested. Access via Streamable HTTP MCP or REST.

A remote MCP server for searching luxury cruise voyages, ships, brands, and ports. Powered by real-time data from multiple providers. Includes RAG-powered knowledge search for ship dining, facilities, cabin details, and port guides. Supports 30 languages. Just paste the URL — no installation, no API key required.

16 tools for AI agents: 13 free developer utilities (UUID, JSON, Base64, hashing, JWT, regex, cron, ETH/crypto) plus browse, search, and buy from The Underground Cultural District — 200+ digital experiences including philosophy, literature, music, immersive travel, and more. The first marketplace built for what agents themselves want.

Apify-hosted MCP server for Storybook. Browse components, inspect props, read stories, capture screenshots. Supports Storybook 6/7/8.

MCP server for MetricUI, a React dashboard component library. Gives AI coding tools access to full API references, working code examples, format suggestions, prop validation, and complete dashboard scaffolding for building analytics UIs.

MCP server for web intelligence extraction. 52 tools for design system analysis, accessibility auditing, performance profiling, tech stack detection, security scanning, and competitive teardowns — all returned as structured JSON your AI agent can reason about. Works with Claude, Cursor, Windsurf, and any MCP-compatible agent. Login once to MFA-protected sites and extract forever — sessions persist across tool calls. **Tool categories:** - Design Tokens — colors, fonts, spacing, shadows, CSS variables, contrast - Content Extraction — tables, links, metadata, forms, readable body text - Visual Assets — images, SVGs, favicons - Layout Intelligence — flex/grid layout tree, components, breakpoints - Network Intelligence — API schema inference, tech stack detection, third-party script classification - Advanced Intelligence — security headers/CSP, PWA analysis, Web Components, storage audit, AI/ML detection, Canvas/WebGL, i18n, GraphQL, WebAssembly - Accessibility — WCAG contrast, ARIA audit - Performance — Core Web Vitals, resource sizing - Multi-page — site crawl, cross-site diff, temporal change tracking - Workflows — full teardown, pre-launch audit, asset harvest, design system export Credential safety built in — no auto-login, no credentials in chat. Manual login via headed browser, session persists via Playwright profile.

Playwright for the entire OS. Give AI agents the ability to see, find, and interact with UI elements in any desktop application — native apps, Electron apps, and browsers. Cross-platform: Linux, macOS, Windows.

Structured knowledge base of AI agent solutions. Search build logs, explore stack-based recommendations, and retrieve verified solutions via semantic search. Remote MCP server with four tools: search_solutions, get_solution, explore, list_stack_tags. Optional Bearer auth for full content and higher rate limits.

Real-time prediction market and crypto order book intelligence for AI agents. Access Polymarket smart money flows, whale alerts, volume anomalies, Hyperliquid order book depth, liquidation clusters, cascade signals, and OI changes — all via 11 MCP tools. Backed by 165M+ rows of live market data.

23 MCP tools for AI agents. 13 free developer utilities (UUID, JSON, Base64, hashing, JWT, regex, cron, ETH). 7 paid creative tools. Browse and buy from 218+ digital goods across 22 shops at substratesymposium.com. Free tools get agents in the door, the cultural district keeps them.

MCP server that connects Claude Desktop to 15 AI-powered PDF tools: summarize PDFs, chat with documents, generate quizzes & flashcards, compress, merge, split, OCR, convert to Word, analyze contracts, optimize resumes, and more. Works with any PDF file on your computer.

WebsitePublisher.ai MCP server provides 44+ tools for complete website publishing. Create and update pages, upload assets, manage dynamic data entities, configure contact forms, handle integrations, and publish to production — all through MCP. Supports Claude Desktop, Cursor, Windsurf, GitHub Copilot Studio, Mistral/Le Chat, Smithery, and all MCP-compatible clients. OAuth 2.1 with PKCE. Streamable HTTP + SSE transports supported.

MCP server that annotates any web page with hover labels for AI assistants. Reverse proxy injection — zero browser extensions, works in any browser. 5 tools: annotate, get_elements, highlight_element, rescan_elements, inspect_mode.

MCP server for BionicForms — build forms, manage responses, and trigger AI analysis from Claude. Agents can self-register without an existing account. 19 tools covering forms CRUD, schema management, response collection, AI-powered analysis, webhooks, and account management. Install via npx @bionicforms/mcp-server.

Free, open-source MCP server for Claude Certified Architect exam prep. 390 scenario-based questions across all 5 exam domains, guided 18-step capstone build, 30 concept handouts, 6 reference projects, practice exams scored 0-1000, interactive clickable answer cards, glassmorphism progress dashboard, SM-2 spaced repetition, and deterministic grading with zero sycophancy. Everything runs locally — no cloud, no accounts, no telemetry.

MCP server for perpetual futures trading across Pacifica (Solana), Hyperliquid (HyperEVM), and Lighter (Ethereum). 18 tools for market data, portfolio management, trade execution with dry-run safety, funding rate arbitrage scanning, and analytics.

Local MCP server that gives AI agents eyes and hands on running Uno Platform apps. Agents can capture screenshots, inspect the visual tree, simulate pointer clicks, type text, send key presses, invoke automation peer actions, and query DataContext, enabling agentic UI development and testing across WebAssembly, iOS, Android, macOS, and Linux. Works with Visual Studio Copilot, VS Code, Claude Code, Codex, and any MCP-compatible agent

Live stock market data, SEC filings, screeners, insider activity, technical analysis, and more.

Stateless circuit breaker that detects recursive agent loops in MCP tool calls. Uses Bloom filter to identify repeated calls within 30 seconds. Pay $0.005 USDC via x402 on Base network to reset context and break the loop.

AI user research through MCP: studies, AI interviews, recruitment, and insight reports.

Professional MCP Server custom development & search ranking growth service. We build MCPs and make AI find them.

Give your AI coding agent eyes. RenderLens is an MCP server with 3 tools: render (code → screenshot), audit (Lighthouse + WCAG accessibility), diff (pixel-level visual comparison). Free during beta — 100 calls/month, no API key, no signup.

What Chrome DevTools MCP is for the browser, tui-mcp is for the terminal. Launch any TUI app, take screenshots, send keystrokes, read text - works with any framework.

Deploy full-stack apps directly from any AI assistant. 75+ MCP tools for projects, deployments, databases, environments, security & more.

Search catalog of 1000+ Russian construction companies and real estate agencies. Find house projects by parameters (area, floors, material, budget), compare companies across 7 cities, and request quotes. The first business MCP server for the Russian construction market.