MCP Servers

Claude Desktop job search assistant: searches LinkedIn plus 8 ATS boards (Greenhouse, Lever, Ashby, Workday, SmartRecruiters, Hacker News, RemoteOK, Remotive), scores each job 0-100 for fit (Claude judges directly, no LLM API key), and shows a ranked, triageable board inline in the chat.

AI job-hunting agent that scans 130+ company careers pages nightly, scores every role 0–100 against your resume with an LLM, sends top matches to Telegram, and drafts tailored cover letters + resume bullets on command. Pluggable free LLM backends (OpenRouter free tier, Claude Code CLI, or Anthropic API). Exposes scanjobs, draftapplication, and exportjobs. Runs locally; does not auto-submit applications.

Associative key-graph memory for LLM agents — recall by association, not vector similarity.

Selects the minimum code context an LLM needs to understand a git diff: graph-based fragment selection under a token budget, deterministic, 30+ languages. Published on PyPI and in the MCP Registry.

Give Your Agent A Voice: x402 pay-per-call speech on Base with 20 voices, 10 personas, 31 languages, granular speed/quality controls, OpenAI-shaped requests, voice samples, and batch audio.

Local-first, zero-egress security scanner for AI-generated / "vibe-coded" JS/TS. Bundles Opengrep, Gitleaks & Trivy behind one CWE-keyed schema and adds AI-code-specific checks (client-side secret exposure, Supabase RLS, prompt-injection & LLM-output XSS sinks). No account, no telemetry.

Cryptographic receipts for AI outputs: notarize any model inference with a signed Ed25519 attestation, sha256 content hash, and Merkle chain-anchor on Base or Solana. $0.001 per call via x402 USDC micropayments; verification is free and needs no wallet. Notarizes the hash, never your prompts. npx -y @forgemeshlabs/x402-notary-mcp

Deterministic AI audit layer: evaluates LLM/agent outputs against configurable policies (jailbreak, safety, quality, wallet, trade, MEV compliance) and writes every verdict to a tamper-evident SHA-256 hash chain — metadata only, never raw content. 17 tools including a full crypto-trading compliance suite. Pay-per-call via x402 (USDC on Base), from $0.01. mcp.fronesislabs.com/mcp

Check whether anything you depend on is known malware, before an agent installs it. checklockfile takes a package-lock.json, yarn.lock or pnpm-lock.yaml and matches every pinned version against published malicious-package advisories in one call, free and with no API key, catching compromised releases like chalk@5.6.1 while leaving their clean releases alone. scanartifact adds deterministic behavioral analysis (no LLM in the serving path) for credential theft, exfiltration, obfuscation, prompt in

Self-hosted web research for MCP agents: search (SearXNG, with DuckDuckGo fallback), crawl, dense+BM25 rerank, and dedupe into a source-grounded, cited prompt. Local ONNX embeddings by default, or bring an OpenAI-compatible embedding API.

Verified SaaS, AI, and LLM pricing for 490+ tools: plans, hidden costs, alternatives, and comparisons. Free, no API key.

Image generation and editing plus video generation (Veo, Sora, Seedance) via OpenRouter or any OpenAI-compatible API. Saves files to disk with token-cheap inline previews, resumable video jobs, live model listing, and a config-diagnostics tool. npx -y mediamcp.

Turn narrated screen recordings (or any local video/audio) into agent-ready data: local Whisper transcript, scene keyframes, OCR, full-text search, and wall-clock anchoring that maps spoken remarks to real timestamps for log correlation. No cloud, no LLM inside. uvx talkthrough-mcp.

Local-first LLM API cost tracker. Captures usage across Anthropic, OpenAI, Qwen, and DeepSeek into a local SQLite ledger and exposes spend queries, provider comparison, and recommendations as MCP tools — with first-class Chinese-provider support (CNY→USD). Install: uvx llm-usage-mcp.

Query your Spanlens LLM observability from any MCP client. 7 read tools for request logs, agent traces, cost stats, anomalies, model-savings, and per-user analytics across OpenAI, Anthropic, and Gemini. Open source, self-hostable. npx -y @spanlens/mcp-server

🎖️ - Query Spanlens LLM observability from Claude Desktop, Cursor, or Continue: request logs, costs, agent traces, anomalies, and user analytics with a read-only API key.

Observability for AI agents: 87 tools to query LLM cost, errors, and latency across OpenAI/Anthropic/Gemini/Mistral, and operate alerts, budget gates, evals, and safety checks from Claude/Cursor. Install: npx -y @argosvix/mcp-server.

🎖️ - Hosted MCP server for LLM Pulse AI visibility analytics, including brand mentions, citations, sentiment, share of voice, recommendations, GEO Writer, Search Console, and AI traffic via a Streamable HTTP endpoint.

Local-first memory engine for AI-agent teams: private/team/project ACL (a hard gate before ranking), associative recall, and federated sync where revocation propagates across nodes. SQLite + FTS5, no LLM required; MCP server + web console included. pip install 'agent-memory-os[mcp]'

Local-first memory for AI agents: a single self-contained Rust binary over one SQLite file. Hybrid BM25/FTS + on-device ONNX embeddings + cross-encoder rerank; abstains instead of returning a confident wrong memory. No server, no API key, zero marginal cost. memkeeper mcp

Compressed Knowledge Graphs (pre-structured dependency DAGs) as MCP context — agents traverse declared edges instead of inferring from text. Open CKG Benchmark (64 domains, 10,838 queries): 3.8× RAG's F1 at 11× fewer tokens, ~42× RDS, 0 fabricated edges by construction. 97 domains, 4 tools, no DB/embeddings. pip install ckg-mcp

Local-first memory for long-horizon AI agents with no LLM at ingest — $0 per message, zero data egress, verbatim source-traceable recall. Typed belief revision, selective forgetting, and a provenance guard that blocks memory-justified destructive actions unless user-confirmed. One midas init wires Claude Code, Cursor, Codex, and 6 more clients to one shared SQLite memory; local web inspector and hash-chained audit log. Benchmarked on LongMemEval/LoCoMo/BEAM with one-command repros — including pu

Persistent memory for AI agents through a simple key-value HTTP API. No vector database or embeddings required. Store, retrieve, search, and share memory across agents with shared namespaces and TTL support. npx -y agentram-mcp

Local-first developer activity aggregator. Indexes commits, PRs, Jira/Linear tickets, Confluence docs, Slack threads, and Calendar events into SQLite + FTS5 + on-device ONNX embeddings; exposes 15 tools, 3 resources, and 3 prompts so Claude Code, Cursor, and Codex can search and cite your past work. brew install --cask pavelpilyak/devrecall/devrecall