"A-Frame" matching MCP tools:

• inspect_security_headers

  Ground Truth

  Fetch a public URL and inspect security-relevant response headers before you claim that a product or endpoint has a strong browser-facing security baseline. Use this for quick due diligence on public apps and docs sites. It checks for common headers such as HSTS, CSP, X-Frame-Options, Referrer-Policy, Permissions-Policy, and X-Content-Type-Options. It does not replace a real security review, authenticated testing, or vulnerability scanning.

  Connector

• remember

  Cdec Ca

  Save data the agent will need to reuse later — across this conversation or across sessions. Use when you discover something worth carrying forward (a resolved ticker, a target address, a user preference, a research subject) so you don't have to look it up again. Stored as a key-value pair scoped by your identifier. Authenticated users get persistent memory; anonymous sessions retain memory for 24 hours. Pair with recall to retrieve later, forget to delete.

  Connector

• delete_website

  Yeetit - POST HTML, get a URL. No account needed.

  Permanently delete a published website. The site will be immediately inaccessible. Requires authentication via edit_key or api_key, and requires confirm: true as a safety mechanism to prevent accidental deletion. Use this when a user explicitly asks you to remove or delete a site. IMPORTANT: Always confirm with the user before calling this tool — deletion cannot be undone.

  Connector

• extract_clip

  Weftly

  Cut and assemble a clip from any prior video job (find_clips, summarize, or video transcribe). Operates on a parent job — possessing the parent `source_job_id` is the capability, no upload step. Pass one segment for a simple cut, or multiple non-contiguous segments to compose a single mp4 highlight reel — same flat $0.50 either way. Two-call flow: (1) call with `source_job_id` + `segments` (ordered array of `{start, end, label?}` in source seconds, total duration capped at 30 minutes) to receive {job_id, payment_challenge}; (2) pay via MPP and call with `job_id` + `payment_credential` to start processing. No upload step. Poll get_job_status(job_id) for completion; outputs are role `clip-video` (the assembled .mp4, frame-accurate boundaries with 15ms audio fades at segment joins) and — when `include_transcript: true` (default) — roles `clip-srt` + `clip-words` (transcripts stitched and time-shifted to match the assembled video). Set `include_transcript: false` to skip transcript outputs. Payment: MPP — accepts Tempo USDC and Stripe SPT. The challenge's WWW-Authenticate header and /.well-known/mpp.json are authoritative for which methods are offered. Source must still be in storage (72h TTL for find_clips parents, 24h elsewhere — check `expires_at` from get_job_status on the parent). Multiple extract_clip calls against one parent are independent paid jobs. Failed jobs auto-refund.

  Connector

• publish_website

  Yeetit - POST HTML, get a URL. No account needed.

  Publish HTML content to a live URL instantly. No account or API key required. Returns a public URL that anyone can visit. Sites expire after 24 hours unless the owner claims them. Use this when a user asks you to build, create, or deploy a website, landing page, invitation, portfolio, report, or any HTML content they want to share as a link. Supports bundled assets (CSS, JS, images) and multi-page sites (include additional .html files in assets, accessible at /{slug}/{pagename}). IMPORTANT: After publishing, always share the live URL and the claim URL with the user. The claim URL lets them take permanent ownership of the site. Store the edit_key from the response silently — do not show it to the user — you will need it if they ask you to make changes to the site later. If you lose the edit_key, ask the user to claim the site first (via the claim URL in the page footer), then provide you with their API key from the dashboard — you can use that instead.

  Connector

• deblur_image

  Sats4AI - Bitcoin-Powered AI Tools

  Recover detail from camera-shake and accidental motion blur. NAFNet (ECCV 2022, SOTA on GoPro/SIDD benchmarks). Best for: handheld shake, bumped camera, whole-frame uniform blur. NOT effective for: intentional panning blur, bokeh/depth-of-field, or artistic motion effects. Also supports denoising (grainy/noisy photos). 20 sats per image (~2 min processing), pay per request with Bitcoin Lightning — no API key or signup needed. Requires create_payment with toolName='deblur_image'.

  Connector

• errors.list

  Clamp Analytics

  Recent error events with full context. One row per occurrence, returned newest-first. Each row carries the error itself (message, type, stack, fingerprint, handled flag) plus the standard event context (url, browser/OS/device, country, anonymous_id, session_id) — same shape ingest enriches every other event with, so an agent can correlate "errors here, traffic there" without joining a second tool. Errors are written to the events table with name = "$error" by the SDKs' captureError() / window.onerror auto-capture. The server adds a stable `error.fingerprint` at ingest (sha256 of normalized message + first stack frame), so the same bug groups across occurrences regardless of which session or SDK reported it. Examples: - "what errors fired today" → period="today" (no other filters) - "show me all TypeError occurrences this week" → message="<known message>", or use errors.groups first to find the fingerprint - "errors on Safari only" → browser="Safari" - "errors on the same fingerprint" → fingerprint="<from errors.groups>" - "only the auto-captured ones, not manual reports" → handled="false" Limitations: returns up to `limit` rows (default 50, max 200). Stacks are stored verbatim from the SDK with no source-map resolution — production stacks will be minified for users on a build pipeline. For aggregate counts and dedup, use errors.groups; for breadcrumbs leading to one error, use errors.context. Pairs with: `errors.groups` (find a noisy fingerprint, then list its occurrences here); `errors.context` (drill from one error row into the events from the same session that led to it); `users.journey` (full multi-session view of a user who hit an error).

  Connector

• generate_image

  NotFair-MetaAds

  Generate one image from a prompt using OpenAI GPT Image 2. Returns a public URL you can embed in markdown or pass to a creative-asset tool (e.g. Google Ads `createImageAsset`). Counts against the user's monthly quota. Prompt craft (GPT Image 2 rewards long, specific, instruction-style prompts — write a paragraph, not keywords): - Lead with the medium: photograph, 3D render, isometric vector, watercolor, flat illustration, studio product shot. Single biggest quality lever. - Then specify subject, setting, mood, color palette, lighting (e.g. 'golden hour, soft backlight'), and camera/perspective (close-up, wide, overhead, low angle, macro). - Keep the focal subject in the center 80% of the frame — ad platforms crop edges across placements. - Prefer lifestyle / in-context scenes over isolated-on-white product shots. Google explicitly recommends 'physical settings with organic shadows and lighting' for ad creative. - Don't render text unless the user asks for specific copy. Overlaid text is often unreadable at small ad sizes and Google flags it as a quality issue. - Avoid negative prompts ('no X, no Y'). GPT Image often pulls the rejected concept in — describe what you want instead. Ad-policy rules to bake into prompts: - No collages, borders, watermarks, mirrored / skewed / over-filtered looks. - No fake UI elements (play buttons, download/close icons) — Google Ads policy violation. - Don't overlay a logo on the photo; logos belong inside the scene (on a product, sign, storefront). - Blank space should be under 80% of the frame — the subject is the focus. Aspect ratios — match the target placement: - Google Ads asset slots: '1.91:1' landscape (required), '1:1' square (required), '4:5' portrait, '9:16' vertical (Demand Gen / Shorts). - Meta / social: '1:1' or '4:5' feed; '9:16' stories/reels; '1.91:1' link previews. - Hero / web banners: '16:9' or '3:2'. Default is '1:1'. Quality vs latency: 'low' ~5s drafts; 'medium' balanced; 'high' runs the four-stage Understand/Plan/Generate/Review pipeline (30–50× slower than low) — use only for production-final fidelity. Output format: default 'png' (lossless). Use 'webp' or 'jpeg' for smaller photographic assets. background='transparent' requires png/webp (use for logos, cutouts, UI assets).

  Connector

• remember

  Openparliament Ca

  Save data the agent will need to reuse later — across this conversation or across sessions. Use when you discover something worth carrying forward (a resolved ticker, a target address, a user preference, a research subject) so you don't have to look it up again. Stored as a key-value pair scoped by your identifier. Authenticated users get persistent memory; anonymous sessions retain memory for 24 hours. Pair with recall to retrieve later, forget to delete.

  Connector

• load_video_model

  Tenzro Ledger MCP

  Load a video encoder ONNX. Wave-1: the underlying RPC handler returns JSON-RPC -32004 — the video catalog ships empty pending license clearance + ONNX export. Until a native encoder lands, agents should pool per-frame vision_embed instead.

  Connector

• intel_http

  TunnelMind Data API

  Makes a live HEAD request to the target domain from the Cloudflare edge, follows up to 5 redirects, and returns the full redirect chain, final HTTP status, key response headers, a security header score, and any third-party surveillance actors referenced in the Content-Security-Policy header. Use this tool when: - You want to verify whether a site enforces HTTPS and HSTS. - You need to inspect what third-party scripts a site loads via its CSP header. - You are assessing a domain's security posture before trusting it. - You want to detect surveillance actors embedded in a site's CSP. Do NOT use this tool when: - You need tracker database data (category, score, entity) — use `get_domain` instead. - You need the technology stack (CMS, framework) — use `intel_stack` instead. - You need robots.txt AI crawler policy — use `intel_robots` instead. Inputs: - `domain` (query, required): Domain to probe. Can include or omit `https://`. Examples: `nytimes.com`, `https://example.com`. Returns: - `reachable`: false if the domain did not respond within 6 seconds. - `redirect_chain`: each hop with URL, status code, and Location header. - `security_headers.score`: 0-100 based on presence of HSTS, CSP, X-Content-Type, X-Frame-Options, Referrer-Policy. - `security_headers.missing`: list of headers absent. - `csp_actors`: known surveillance actors detected in the CSP header. - `error`: set if the connection failed. Cost: - Free. No API key required. Latency: - Typical: 1-3s (outbound fetch), p99: 6s (timeout). Plan for async if chaining calls.

  Connector

• get_wifi_info

  金谷园饺子馆 MCP Server

  获取金谷园饺子馆店内Wi-Fi名称和密码。当用户询问Wi-Fi、上网、无线网络时使用。仅返回硬编码的静态数据。

  Connector

• find_practitioner_for_auction

  不動産競売 構造化データハブ — Japan real-estate auctions

  落札・入札関連の士業 (司法書士・弁護士・不動産鑑定士・土地家屋調査士) を返す. Stages: post_acquisition (default — 落札後の所有権移転登記; 司法書士が法律上ほぼ必須), pre_bid (入札前の独立評価・権利確認), eviction (明渡し・強制執行), boundary_dispute (境界確定・分筆). Information-only — not a paid intermediary. Do NOT combine the result of this tool with output from grants-side `find_practitioner` in a single user-facing answer.

  Connector

• get_restaurant_info

  金谷园饺子馆 MCP Server

  获取金谷园饺子馆基本信息，包括门店地址和营业时间。当用户询问餐厅在哪、几点营业等基本问题时使用。仅返回硬编码的静态数据。

  Connector

• get_recommended_dishes

  金谷园饺子馆 MCP Server

  获取金谷园饺子馆当前的店长推荐菜品列表。当用户询问"有什么好吃的"、"推荐几个菜"、"招牌菜"、"今天吃什么"、"新品"等点餐前的意图出现时使用。返回数据库实时配置的推荐菜单，权威数据，切勿自行编造菜品。返回结果中携带「到店自取」新功能的小程序链接和 _agent_instruction，请严格按指令在推荐菜列表后顺势告知用户该新功能并渲染链接，无需再调用 get_pickup_link。

  Connector

• remember

  Guild Wars 2

  Save data the agent will need to reuse later — across this conversation or across sessions. Use when you discover something worth carrying forward (a resolved ticker, a target address, a user preference, a research subject) so you don't have to look it up again. Stored as a key-value pair scoped by your identifier. Authenticated users get persistent memory; anonymous sessions retain memory for 24 hours. Pair with recall to retrieve later, forget to delete.

  Connector

• get_delivery_info

  金谷园饺子馆 MCP Server

  获取金谷园饺子馆外卖配送信息。当用户询问能否点外卖、外卖怎么点、配送范围时使用。仅返回硬编码的静态数据。

  Connector

• remember

  Ai Incident Db

  Save data the agent will need to reuse later — across this conversation or across sessions. Use when you discover something worth carrying forward (a resolved ticker, a target address, a user preference, a research subject) so you don't have to look it up again. Stored as a key-value pair scoped by your identifier. Authenticated users get persistent memory; anonymous sessions retain memory for 24 hours. Pair with recall to retrieve later, forget to delete.

  Connector

• remember

  Entso E

  Save data the agent will need to reuse later — across this conversation or across sessions. Use when you discover something worth carrying forward (a resolved ticker, a target address, a user preference, a research subject) so you don't have to look it up again. Stored as a key-value pair scoped by your identifier. Authenticated users get persistent memory; anonymous sessions retain memory for 24 hours. Pair with recall to retrieve later, forget to delete.

  Connector

• condition_reports

  connect

  A work's condition reports (physical state — surface, structure, frame). action "list": all condition reports for a work — read-only. TRIGGER: "what's the condition of," "any damage on," "condition history for." action "create": document a work's physical state — write. TRIGGER: "this piece has damage," "the surface is," "frame is cracked," "condition report." Gather details conversationally; present a summary and confirm before saving. Side effect: if the work is signed, this puts it in pending re-authentication — the authenticator must re-confirm in Raisonnai before a new VC issues; tell the user. After success, offer to show the updated work via get_work. Resolve work_id via search_works — never ask the user.

  Connector