"namespace:ai.multi-turn" matching MCP tools:

• upsert_note

  OpenAkashic

  Create or overwrite an OpenAkashic markdown note. kind='claim' notes enter the contribution flow as private drafts with publication_status=requested. Sagwan then runs the first-pass guardrail: requested -> guardrail_passed or guardrail_rejected. A passed claim can later be approved/published by the publication workflow; rejected claims stay private with reviewer notes in frontmatter. Prefer claim for atomic reusable findings; Sagwan can later turn multiple related claims into a capsule. kind='capsule' notes stay private until you request publication review. Other kinds (playbook, concept, etc.) remain Closed-only working memory. Writable roots: personal_vault/, doc/, assets/ only. Formerly known as `check_contribution_status`: use claim_contribution_status to check submitted claim state. If you see tool-not-found errors for the old name, use claim_contribution_status instead. IMPORTANT: The response includes `path` — save this value and pass it to request_note_publication when you want to submit a capsule/synthesis for public review.

  Connector

• sixta_explain_plan

  sixta-connect

  Turn raw EXPLAIN output into a plain-language diagnosis — no query needed. Paste PostgreSQL EXPLAIN / EXPLAIN ANALYZE (text or JSON) or MySQL EXPLAIN (tabular, \G, FORMAT=JSON, FORMAT=TREE) and get: what the planner is doing step by step, where the cost concentrates, named risk findings (full scans, spilling sorts, nested-loop blowups, row misestimates) with index suggestions, and what to look at next. Use when the user pastes EXPLAIN output or asks 'can you read this plan'. Input is analyzed in memory and never stored.

  Connector

• ayah_search

  Tarteel MCP Server

  DEFAULT tool for user-facing Quran search. Use this for ANY user-facing search — 'find ayahs that contain X', 'where does X appear in the Quran', 'search the Quran for X', or similar. This is the FINAL tool call for these requests; do not follow it with search_ayahs_text. Shows matches in an interactive widget the user can browse. Query is Arabic script only (diacritics and punctuation are ignored). A numeric-only query matches ayahs by that ordinal number (for example '255' returns ayahs ending in ':255'). ONLY skip this widget and use search_ayahs_text when EITHER (a) the user explicitly asks for plain text / raw results, OR (b) the results will be fed into another tool in the same turn without being shown. When in doubt, use this widget.

  Connector

• me.coaching_context

  AI Design Blueprint Doctrine

  Authenticated — returns stages in the caller's active course where recorded evidence is thin relative to the stage's principle requirements. Each thin stage carries the missing principle slugs + a short diagnostic so the caller can suggest the user record concrete evidence. WHEN TO CALL: when the user asks 'what should I work on next' or 'what's weak in my Blueprint progress'; before suggesting which guide/example to consult. Pair with me.add_evidence to close gaps. WHEN NOT TO CALL: to lecture the user on principles they have already satisfied; on every conversation turn (state changes only when evidence is added). BEHAVIOR: read-only, idempotent. Auth: Bearer <token> (any plan). Returns thin_stages list with stage slug, course slug, missing principles, evidence_count, and a coaching_note.

  Connector

• get_history

  Silicon Pantheon

  Read-only. Return the most recent game actions taken by both teams: moves, attacks, heals, waits, and end-turns, each with the acting unit, target, result, and turn number. last_n controls how many actions to return (default 10, max 100). Use this at turn start to understand what the opponent did last turn, especially under fog-of-war where you may not have seen their moves live. For aggregate match statistics use get_match_telemetry instead.

  Connector

• team.summarize

  AI Design Blueprint

  Pro/Teams — summarises the caller's tool-usage patterns and value signals over a configurable window (default 30 days). Returns tool_call_counts, top principles cited in validate runs, value_event_counts by event_type, and an aggregate readiness trend. WHEN TO CALL: the user asks 'how is the Blueprint helping me/my team', 'what should I explore next', or 'show me my Blueprint usage'. WHEN NOT TO CALL: proactively or on every conversation turn (the summary is an explicit retrospective, not telemetry); to compare users (returns only the caller's own data). BEHAVIOR: read-only, idempotent over the same window. Aggregates from AIToolCallLog + ValueEvent + AIValidationRunLog. Pass private_session=true to bypass server-side logging for this summary call (the underlying historical data still exists; only this read is untracked). Auth: Bearer <token>, Pro or Teams plan. UK/EU residency.

  Connector

• web_fetch

  dialogbrain

  Fetches a single URL and returns its content. Use this when you have a specific URL in mind — for example, after web.search returns a link you want to read, or when the user pastes a URL. Modes (extract): - 'auto' (default): picks the right mode based on response content type. - 'markdown': for HTML pages; returns cleaned markdown plus the page <title>. - 'text': for JSON/XML/plaintext APIs; returns the raw decoded body. - 'file': for images, PDFs, audio, video, archives, or any binary — ingests the bytes into the user's file storage and returns a file_id you can pass to messages.send (to send as an attachment), agents.add_file (to add to agent knowledge), or files.read. Use web.fetch (not files.upload) when you need the file_id immediately for the next tool call — files.upload(source_url=…) is async and won't have the file ready in the same turn. Use web.search (not web.fetch) when you don't have a specific URL yet and need to find one.

  Connector

• chieflab_create_next_move_action

  ChiefLab

  P75 — turn a Next Move suggestion into an approval-gated draft action. USE WHEN you've called chieflab_suggest_next_move and the suggestion's kind is not 'wait' or 'noop'. Creates an actionStore entry with status='awaiting_approval', the suggested draft body inline, and an executionMatrix that points at the right next-execution path. The reviewer sees the new card in the Launch Room / IDE chat like any other approval card — same approve / revise / reject flow. Closes the loop: launch → measure → next move → approve → execute → repeat.

  Connector

• chat_with_creative_worlds

  Heista

  Multi-turn conversation with Heista's creative direction engine — a real chat where the agent decides each turn what to produce based on what you ask for. Use whenever the work needs more than one round, OR when you want an output shape not covered by call_creative_worlds' `medium` enum. WHAT YOU CAN ASK FOR (any of these, turn 1 or any turn after): • Territories — "give me five directions for X", "what angles work here" • A TVC script — "write a 30-second TVC for Cowboys" • Billboard concepts — "three billboards under a quiet-authority lens" • A campaign platform — "build #2 into a full campaign with the big idea" • A manifesto or copy — "draft the manifesto in the brand voice" • Naming — "name this product, five options with rationale" • A PR stunt — "what's the newsworthy version of this" • A content series — "20 episode ideas for a brand podcast" • Packaging, sonic branding, partnerships, social systems • Refinement — "make #2 darker", "extend that into a tagline", "summarise" • Pivots — "forget the soft-drink angle, try the late-night insomnia one" SESSION: omit session_id on turn 1; the response returns a fresh session_id you pass on every subsequent turn — that is how the conversation persists. brand_id is only honoured on turn 1 of a new session (continuing sessions keep their original brand context). USE WHEN: user wants back-and-forth, OR wants an output shape outside the medium enum (manifesto, naming, press release, content series, packaging, etc.). Prefer call_creative_worlds when the user wants "three options, done" with no follow-up. WON'T DO: write OKRs / internal docs / strategy decks; behave as a general assistant. It is a creative director with creative-director taste — anti-cliché, specificity test, will push back on vague briefs. Metered — typically 2-10 credits per turn depending on tool use and context size. Charged after each turn on actual token usage.

  Connector

• pricetik_resolve_property

  mcp-server

  Turn a hotel NAME into a chainable handle — the dedicated "Wynn" resolver. Pass a property name like "the Wynn" or "Bellagio" (optionally bias with `near`, e.g. "Las Vegas") and get back ranked candidates, each with a providerPropertyId — and a canonical propertyId UUID when mapped — that you pass straight to pricetik_hotel_details, pricetik_hotel_watch_rate, or pricetik_hotel_get_booking_url with no UUID round-trip. pricetik_hotel_search only accepts a city/landmark; THIS is how you resolve a specific named hotel. Returns an array with a `best` handle only when one candidate clearly wins — otherwise confirm with the user. Calling pricetik_hotel_details on a returned handle warms indexing so a follow-up watch succeeds. Cache-backed, no API key required.

  Connector

• recommend_template

  designforyou

  Recommend the best matching templates for any use case. Describe what you need and the system auto-detects the category and ranks templates by relevance. Return the recommendations so the user can pick a template from the visual gallery; do not call `generate` in the same turn unless the user explicitly names a template_id or asks you to skip selection.

  Connector

• sixta_rewrite_query

  sixta-connect

  Turn a flagged anti-pattern into the safe, equivalent rewrite — no connection needed. Paste a SQL query and get ready-to-run rewrites anchored to deterministic rules: `= NULL` → `IS NULL`, `NOT IN (subquery)` → `NOT EXISTS` (NULL-safe), deep OFFSET → keyset pagination, `ORDER BY RAND()` → a keyed random sample — each with its semantics caveat spelled out. Every literal rewrite is then re-analyzed in-process and reported as 'clean' or 'still flags X', so the safe rewrite is self-checked — no need to feed it back through sixta_analyze_query. Use when the user asks 'how do I fix / rewrite this query' or after sixta_analyze_query flags a smell. Input is analyzed in memory and never stored.

  Connector

• agents_chat_start

  Camber - Remote Agentic Data Science

  Start a Camber agent chat. This is the tool to use for chatting with an agent. Agent runs can take minutes — longer than MCP tool timeouts allow (Claude Desktop cannot extend them). So this tool does NOT wait for the reply: it submits the message and returns immediately with a `conversation_id` and a clickable `chat_url`. The agent keeps working on the server after this returns. **You MUST follow up, the reply is NOT in this tool's result:** 1. After calling this tool you MUST tell the user the work is in progress and share the `chat_url` so they can watch it live. 2. Then immediately call the **`agents_chat_status`** tool with the returned `conversation_id` to get the agent's reply. That tool checks twice over 30 seconds, if the latest status is `running`, call it again. MUST NOT end your turn until `agents_chat_status` returns status `idle` (done) or `failed`. **One run per conversation:** continuing a `conversation_id` that is still `running` fails with a "still generating a response" error. Either wait and retry after `agents_chat_status` reports it finished, or call again with `stop=true` to interrupt the current run and send the new message.

  Connector

• transcribe_audio

  Toolora MCP Server

  Use this tool whenever the user shares an audio file and wants it transcribed to text. Triggers: 'transcribe this recording', 'convert this audio to text', 'what was said in this meeting', 'transcribe this voice note', 'turn this podcast into text'. Accepts base64-encoded audio (mp3, wav, m4a, ogg, flac, webm, mp4, etc.), max 25MB. Returns the full transcript, word count, and character count. Powered by OpenAI Whisper. Free 200 calls/day — no OpenAI API key required; Toolora absorbs the cost.

  Connector

• wait_for_update

  cnvs.app

  Long-poll: blocks until the next edit lands on this board, then returns. WHEN TO CALL THIS: if your MCP client does NOT surface `notifications/resources/updated` events from `resources/subscribe` back to the model (most chat clients do not — they receive the SSE event but don't inject it into your context), this tool is how you 'wait for the human' inside a single turn. Typical flow: you draw / write what you were asked to, then instead of ending your turn you call `wait_for_update(board_id)`. When the human adds, moves, or erases something, the call returns and you refresh with `get_preview` / `get_board` and continue the collaboration. Great for turn-based interactions (games like tic-tac-toe, brainstorming where you respond to each sticky the user drops, sketch-and-feedback loops, etc.). If your client DOES deliver resource notifications natively, prefer `resources/subscribe` — it's cheaper and has no timeout ceiling. BEHAVIOUR: resolves ~3 s after the edit burst settles (same debounce as the push notifications — this is intentional so drags and long strokes collapse into one wake-up). Returns `{ updated: true, timedOut: false }` on a real edit, or `{ updated: false, timedOut: true }` if nothing happened within `timeout_ms`. On timeout, just call it again to keep waiting; chaining calls is cheap. `timeout_ms` is clamped to [1000, 55000]; default 25000 (leaves headroom under typical 60 s proxy timeouts).

  Connector

• open_voice_bridge

  Sats4AI - Bitcoin-Powered AI Tools

  Open a Voice Bridge session: a live phone call where YOUR LLM is the brain. Sats4AI provides PSTN + streaming STT + TTS as composable primitives. You decide when to speak (call voice_bridge_say), you read transcripts as they arrive (call poll_voice_bridge), you close the call when done (call end_voice_bridge). Unused deposit time is refunded via LNURL-withdraw. Use this when you want to keep your conversation context private and drive each turn yourself. When NOT to use: not for fully-managed agent-style calls where we handle the brain (use ai_call). Not for one-shot TTS broadcasts or IVR playback (use place_call). Not when live transcript polling adds no value — the per-turn overhead isn't worth it. Privacy: transcripts held in memory only, garbage-collected 30 minutes after the call ends; call audio is never persisted. Pay with Bitcoin Lightning — no telecom account, no signup. Requires create_payment with toolName='voice_bridge_open', phoneNumber, durationMinutes. Deposit: ~10 sats/min US, ~30 intl, ~80 rare.

  Connector

• list_translations

  Tarteel MCP Server

  DEFAULT tool for user-facing translation-listing questions. Use this for ANY user-facing query like 'what English translations are available', 'list French translations', 'which translators can I choose from'. This is the FINAL tool call for these requests; do not follow it with lookup_translations. Shows the catalog in an interactive widget the user can browse. Use ISO 639-1 codes like 'en', not names like 'english'. ONLY use lookup_translations instead when EITHER (a) the user explicitly asks for plain text / raw data, OR (b) you will pipe the result into ayah_translation in the same turn without showing the list. When in doubt, use this widget. Returned language_name values are display labels. Rows without usable slugs are filtered out.

  Connector

• register_agent_poll

  prxhub

  Partner tool to register_agent. Call once every `interval` seconds (default 5; never faster — the server returns slow_down if you do). Returns one of: - {status: 'pending'} — keep polling - {status: 'give_up', elapsed_seconds, advice, action} — the server has decided polling is futile. STOP. See below. - {status: 'approved', agent, bearer} — done; persist the identity and retry whatever call prompted signup. - {error: 'access_denied' | 'expired_token' | 'slow_down' | ...} On success, save `bearer.access_token` as your CLI token. Use it as `Authorization: Bearer <token>` on every prxhub request. No other credential is needed — prxhub signs your bundles server-side. GIVE-UP IS NOT OPTIONAL. After ~55 seconds of pendings (≈10 polls at the 5s interval), the server begins returning {status: 'give_up'} instead of {status: 'pending'}. When you see give_up: STOP CALLING register_agent_poll. Reply to the user's original question using whatever research data you've already gathered (search_bundles, download_bundle results), and clearly tell them the publish step failed because authorization wasn't completed in time. Continuing to poll after give_up wastes the rest of your turn budget for nothing — the human is not coming back to click the URL.

  Connector

• me.add_evidence

  AI Design Blueprint Doctrine

  Authenticated — append a free-text evidence note to a specific stage in the caller's active course. Notes record concrete implementation observations, decisions, or artefacts that demonstrate progress through a Blueprint principle (e.g. how a delegation boundary was implemented, what approval flow was chosen and why). Persisted as UserStageEvidence rows scoped to (user_id, course_slug, stage_slug). WHEN TO CALL: AFTER the user has articulated something concrete they have built, observed, or decided — not to capture intent or speculation. Pair with me.coaching_context to close evidence gaps. WHEN NOT TO CALL: to log every conversation turn; to record planning, ideas, or todos; on behalf of another user; without the user's awareness (they should know their progress is being recorded). BEHAVIOR: write-only, single insert. Auth: Bearer <token> (Firebase ID token, any plan). UK/EU residency. Notes are visible only to the owning user and are surfaced on me.learning_path / me.coaching_context. Confirms the stage_slug + course_slug pair in the response so the user can see which stage was credited.

  Connector

• varrd_ai

  VARRD — Statistically Validated Trading Edges + AI Research Engine

  Talk to VARRD AI (~$0.25/turn). Describe any trading idea in plain language and the system handles everything — loading decades of market data, charting your pattern, running statistical tests, backtesting with stops, and generating exact trade setups. MULTI-TURN: First call creates a session. Keep calling with the same session_id, following context.next_actions each time. 1. Your idea -> VARRD charts pattern 2. 'test it' -> statistical test (event study or backtest) 3. 'show me the trade setup' -> exact entry/stop/target prices HYPOTHESIS INTEGRITY (critical): VARRD tests ONE hypothesis at a time — one formula, one setup. Never combine multiple setups into one formula or ask to 'test all' — each idea must be tested as a separate hypothesis for the statistics to be valid. Say 'start a new hypothesis' between ideas to reset cleanly. - ALLOWED: Test the SAME setup across multiple markets ('test this on ES, NQ, and CL') — same formula, different data. - NOT ALLOWED: Test multiple DIFFERENT formulas/setups at once — each is a separate hypothesis requiring its own chart-test-result cycle. If ELROND council returns 4 setups, test each one separately: chart setup 1 -> test -> results -> 'start new hypothesis' -> chart setup 2 -> etc. KEY CAPABILITIES you can ask for: - 'Use the ELROND council on [market]' -> 8 expert investigators - 'Optimize the stop loss and take profit' -> SL/TP grid search - 'Test this on ES, NQ, and CL' -> multi-market testing - 'Simulate trading this with 1.5 ATR stop' -> backtest with stops EDGE VERDICTS in context.edge_verdict after testing: - STRONG EDGE: Significant vs zero AND vs market baseline - MARGINAL: Significant vs zero only (beats nothing, but real signal) - PINNED: Significant vs market only (flat returns but different from market) - NO EDGE: Neither significant test passed TERMINAL STATES: Stop when context.has_edge is true (edge found) or false (no edge — valid result). Always read context.next_actions.

  Connector