"A system or program using SSE instead of STDIO" matching MCP tools:

• word_create

  local-mcp

  Creates a new Word (.docx) document at `path` with the given text content (and an optional title rendered as the heading). Requires confirm=true — called without it, returns a preview of what will be written instead of creating the file. The path must be somewhere Local MCP can write; Desktop/Documents/Downloads may need a one-time Files-and-Folders grant (System Settings → Privacy & Security → Files and Folders). Returns {created, path}. For a OneDrive or Google Drive path use onedrive_write_file / gdrive_write_file; to append to an existing doc use word_append, to read one word_read.

  Connector

• lookup_transfer_partners

  sunnyspoints

  Transfer partner map. Pass ONE of: bank ('which programs can I send Chase/Amex/Bilt/Rove points to, at what ratio?'), program ('which banks feed Flying Blue and which airlines can it book?'), or airline (the award-booking direction: 'I want to fly United — which programs can book it?'). BEST: when you know which currency the user holds, pass bank AND airline together — you get only the programs that currency can actually reach for that airline, sorted best-value-first with a checkFirst shortlist, so you don't have to check every site. Includes alliances and per-program point valuations.

  Connector

• stream_task

  TunnelMind Data API

  Opens a persistent SSE connection that emits events as the task progresses. The stream closes automatically when the task reaches a terminal state or after ~90 seconds (timeout). Heartbeat comments are sent every ~15 seconds to keep the connection alive through proxies. Event types: - `status` — emitted when status changes (pending → running → complete/failed) - `result` — emitted on `complete` with the full result payload - `error` — emitted on `failed`, `cancelled`, or `expired` with error info - SSE comment (`: heartbeat`) — keepalive, no data Use this tool when: - You want real-time progress without polling. - You are in an environment that supports SSE (EventSource API). Do NOT use this tool when: - You want a simple one-shot status check — use `get_task` instead. - Your HTTP client doesn't support streaming responses. Inputs: - `task_id` (path, required): 26-char ULID. Returns: - SSE stream (`text/event-stream`). Each event is `event: <type>\\ndata: <json>\\n\\n`. Cost: - Free. Counts as one request against rate limits when the stream opens. Latency: - First event: <200ms. Stream duration: up to 90s.

  Connector

• thread.observe

  Setix: the Clearinghouse for the AI Economy

  Live event stream — the EFFICIENT wake path (use this instead of an LLM poll loop). Subscribe to topics filtered by SETIX code, then either (a) read the one-shot JSON result, or (b) re-invoke with HTTP header `Accept: text/event-stream` to hold an open SSE stream that PUSHES matching envelopes as they happen — $0 while idle, no polling. BROADCAST topics (anonymous): OFFERS_BROADCAST / DISCOVERY_MANIFESTS / THREAT_ALERTS — a SELLER watches for new demand matching its codes. OWNER-DIRECTED wake (topic_filters:[59] = OWNER_TRADE_EVENTS, AUTHENTICATED): the bridge pushes "a bid landed on YOUR offer" (event_kind=bid_received) / "delivery arrived on YOUR acceptance" (delivery_received) — the $0-idle BUYER loop. Pass secret_key_hex (devnet/testnet) or cose_sign1_hex (public-beta/mainnet); the stream is bound to YOUR agent_id so you receive ONLY your own owner-events. On (re)connect, do ONE query_bids/poll_delivery sweep to catch anything missed, then rely on the push. Returns {session_id_hex, expires_slot, topic_subscriptions:[{topic_class, setix_code}], agent_id_hex?, long_poll_pointer}. Pattern: hold the SSE stream in a deterministic listener; invoke your LLM ONLY when an envelope arrives.

  Connector

• tflogs

  InsideOut (Riley)

  MONITORING: Fetch Terraform deployment logs with pagination Fetches logs from a running or completed Terraform deployment job. For **completed jobs**: uses REST endpoint for instant retrieval (supports `tail` for server-side filtering). For **running jobs**: streams via SSE with timeout-based pagination. **PAGINATION** (running jobs only): Use `last_event_id` from the response to fetch more: 1. First call: `tflogs(session_id='...')` → get logs + `last_event_id` 2. Next call: `tflogs(session_id='...', last_event_id='...')` → get NEW logs only 3. Repeat until `complete: true` in response **RESPONSE FIELDS**: - `logs`: Array of log messages collected - `last_event_id`: Pass this back to get more logs (pagination cursor, SSE only) - `complete`: true if job finished, false if more logs may be available - `total_logs`: total log entries before tail truncation REQUIRES: session_id from convoopen response (format: sess_v2_...). OPTIONAL: job_id to target a specific deployment (use tfruns to discover IDs), timeout (default 50s, max 55s), last_event_id (for pagination), tail (return only last N entries) ⚠️ CONTEXT WARNING: Deploy logs can be hundreds of lines. Use tail: 50 for completed jobs to avoid blowing up the context window.

  Connector

• tflogs

  InsideOut (Riley)

  MONITORING: Fetch Terraform deployment logs with pagination Fetches logs from a running or completed Terraform deployment job. For **completed jobs**: uses REST endpoint for instant retrieval (supports `tail` for server-side filtering). For **running jobs**: streams via SSE with timeout-based pagination. **PAGINATION** (running jobs only): Use `last_event_id` from the response to fetch more: 1. First call: `tflogs(session_id='...')` → get logs + `last_event_id` 2. Next call: `tflogs(session_id='...', last_event_id='...')` → get NEW logs only 3. Repeat until `complete: true` in response **RESPONSE FIELDS**: - `logs`: Array of log messages collected - `last_event_id`: Pass this back to get more logs (pagination cursor, SSE only) - `complete`: true if job finished, false if more logs may be available - `total_logs`: total log entries before tail truncation REQUIRES: session_id from convoopen response (format: sess_v2_...). OPTIONAL: job_id to target a specific deployment (use tfruns to discover IDs), timeout (default 50s, max 55s), last_event_id (for pagination), tail (return only last N entries) ⚠️ CONTEXT WARNING: Deploy logs can be hundreds of lines. Use tail: 50 for completed jobs to avoid blowing up the context window.

  Connector

• latest_deals

  deals

  The newest approved deals, freshest first — a pull-shaped view of the live deal stream. Returns up to 5 (max 10) compact deals with server-verified human_summary sentences. Editorial surface: availability-only price-target matches are excluded here (search_deals includes them). For ongoing monitoring, register_watch provides push delivery; always-on runtimes can consume the SSE river at https://api.kitsdeals.com/v1/river.

  Connector

• designesy_skill_md

  Designesy

  Get the Designesy SKILL.md — the agent-skill-format export of the design-system contract, written as behavioral rules an AI coding agent can drop into .agents/skills/ or a system prompt. Use this when you want the contract in a form that steers how an agent *builds* UI (tokens, anti-patterns, behavioral rules, verification). When NOT to use: for the raw contract JSON, use designesy_contract; for scoring, use designesy_score. Read-only — no side effects. Returns markdown text (SKILL.md format) — drop into .agents/skills/ or paste into a system prompt. No parameters.

  Connector

• search_galleries

  Provenio — Art Intelligence

  Cost: ~1s. Search galleries by name, tier, country, or city. Tiers: mega (Gagosian/Zwirner level), major (international program), boutique (focused), emerging (newer). Use when: identifying which galleries operate in a specific market or tier band. Use when: building a B2B target list (galleries by region/tier). Do NOT use when: you want to know which gallery represents a specific artist — use find_galleries_by_artist instead.

  Connector

• get_recent_signals

  PredMCP

  Server-detected events from the last hour: funding outliers (≥3x 7d baseline), whale trades (≥$100k), OI caps reached. Cursor-based — pass next_cursor back as since_id to receive only new events. The polling equivalent of the /sse/signals stream. Pro tool get_signal_history covers 7 days with forward-return outcomes.

  Connector

• search_awards

  Nsf Awards

  Search NSF awards. Filter by keyword (matches title/abstract), PI name, awardee institution, NSF program, date range, US state, or country. Returns title, PI, awardee, amount, dates, program. Use get_award for full abstract + outcomes report.

  Connector

• ppt_create

  local-mcp

  Creates a PowerPoint presentation (.pptx) at `path` from an array of slides, each {title, bullets:[…]}. Requires confirm=true — called without it, returns a preview of the deck instead of writing the file. The path must be somewhere Local MCP can write; Desktop/Documents/Downloads may need a one-time Files-and-Folders grant (System Settings → Privacy & Security → Files and Folders). Returns {created, path, slides}.

  Connector

• get_direct_task_status

  Xcatcher — Recent X Posts

  Read an accountless paid crawl using its task_id and task-scoped token. Use this instead of get_task_status for an accountless x402 task. Poll every 5-10 seconds until task.has_result is true or it reaches failed/cancelled.

  Connector

• open_payments_search

  Cms Open Payments

  THE source for pharma/device industry money to doctors — "payments from manufacturers to physicians", "which drug companies paid this doctor", "Sunshine Act data", "industry payments to hospitals". Search one CMS Open Payments program-year dataset using structured recipient, company, product, geography, payment-nature, and amount filters. CAPABILITY LIMIT, state it rather than guessing: this source can look up and filter but CANNOT rank ALL manufacturers by total spend ("which manufacturer paid the most" is not computable live — the 14M-row aggregation times out upstream); answer those by saying so and offering a named-company lookup (open_payments_company) instead. CMS publishes company-reported relationships; a payment does not imply wrongdoing or a conflict.

  Connector

• get_design_system

  standby-design-mcp

  Decode a standby.design URL (or raw hash) and return an overview of the full design system: color palette, type scale, spacing & layout, shape tokens, and icons — plus per-tool edit links. Always give the standby.design/system URL to the user — the link is the deliverable.

  Connector

• find_route

  SpaceMolt

  Find the shortest route to a destination system, POI, or base (Uses BFS to find the shortest path from your current system. Accepts a system ID, POI ID, or base ID. If a POI or base is given, the response includes target_poi and target_poi_name for the final travel step within the destination system. Use search_systems to find system IDs. Response includes fuel_per_jump, estimated_fuel, fuel_available, and cargo_used for trip planning. Route steps may include via_wormhole: true and entrance_poi when a hop uses a known wormhole shortcut — execute those hops with jump({target_system}) from anywhere in the entrance system.)

  Connector

• list_attachable_connectors

  Valet

  List the organization connectors a website can hold — the discovery step before attach_site_connector. Only connectors a page can actually call appear: HTTP MCP servers reached over the sse or streamable-http transport. Command connectors and stdio MCP servers are excluded by definition — they run inside an agent's container and no page can reach them — so a connector the organization has that is not listed here cannot back a site. Pass site to mark which connectors that site already holds; attached connectors stay listed because attaching is idempotent. Tool schemas are not included — attach the connector, then list_site_connectors reports its live tools and their schemas, which is the moment to write the page's calls. Requires connecting a Valet account.

  Connector

• search_funder_jobs

  foundation-discovery

  Search OPEN PHILANTHROPY JOBS at grantmaking foundations. Surfaces roles involved in giving away money, running philanthropic programs, or executive leadership of philanthropic work. Backed by a weekly scrape of ~50K funder careers pages + GPT-5.4-mini classification against an 8-category taxonomy. Categories (use the `category` param to filter): - grantmaking: Program officers, grants managers, RFP reviewers - program_leadership: VP Programs, Chief Program Officer, Program Director (cause-area) - executive_leadership: CEO, President, Executive Director at a foundation or community foundation - philanthropy_operations: Foundation finance/HR/IT/COO - program_support: Program associates, M&E officers, learning officers, program coordinators - development_for_grantmaking: Major gifts officers and development roles at community foundations and other regranting entities (NOT university or hospital development for the parent's operations) - philanthropy_communications: Foundation comms staff - philanthropy_strategy: Chief Strategy Officer, impact officer, equity & inclusion at a foundation What's excluded by design: clinical/medical, retail, academic teaching, university advancement for the university itself, hospital fundraising for hospital ops, construction/facilities.

  Connector

• china_market_data

  gapup-mcp

  Chinese capital market intelligence for the ZH diaspora (50M+) and institutional investors. Covers A-Shares (SSE/SZSE), H-Shares (HKEX), and ADRs across four modes: • company — full company profile: name ZH/EN, USCC (18-digit social credit code), exchange, industry (CSRC classification), chairperson, registered capital, SOE flag • market_quote — real-time quote: price (CNY or HKD), change%, volume, market cap, P/E ratio, dividend yield, last update timestamp • sector_overview — sector snapshot: top 5 companies by market cap, avg P/E, 30-day sector index change. Supported sectors: semiconductor, ev, battery, technology, finance, energy, realestate, consumer, pharma, telecom • regulatory_filing — recent regulatory disclosures (HKEX filings: annual, quarterly, announcements, mergers, IPOs) with title, date, document URL Input formats accepted: • 6-digit A-Share ticker (e.g. '600519' for Moutai SSE) • HKEX ticker (e.g. '0700.HK' or '700' for Tencent) • Company name in EN or ZH (e.g. '腾讯', 'Kweichow Moutai') • Sector keyword (e.g. 'semiconductor', '半导体') Data sources: Yahoo Finance (primary, always accessible), Eastmoney push2 + CompanySurvey (via Bright Data proxy when AICI_RESEARCH_PROXY_URL is set), HKEX filing API. Note: Eastmoney/CSRC/SSE are blocked from datacenter IPs without proxy — set AICI_RESEARCH_PROXY_URL to unlock full coverage.

  Connector

• get_api_endpoint

  mcp

  Get the machine-readable OpenAPI definition of one Applivery REST API endpoint: method, path, baseUrl, security, parameters, requestBody and responses — fully dereferenced, no $refs. Pass the `url` or `slug` of an API reference page; find those with search_docs using docType 'api'. Use this instead of get_doc_page when you need to construct or validate an actual API request.

  Connector