This listing identifies a third-party MCP server. MCPBundles is not affiliated with, sponsored by, or endorsed by the service or its operator unless stated on this page.
This server may ask you to sign in when your AI client connects.
Last probed Nov 3, 2026 · mcp.langbot.ai
26tools discovered
For AI agents — MCPBundles Directory API
For MCP client apps (human connect):
MCP connect URL: https://mcp.langbot.ai/mcp
Skill page: https://www.mcpbundles.com/skills/langbot-mcp-635d2e8f8c
Explore capabilities (26)
read 17
other 8
Showing 25 of 26
Showing 25 of 26 tools from the public preview. Connect the server to inspect the full live tool list.
Get10
List5
Search2
Sim2
Subscribe2
Other capabilities4
List Live Events
read
list_live_events
List live and upcoming events Langbot is listening to, with their sources, markets and phrases.
List Live Events — List live and upcoming events Langbot is listening to, with their sources, markets and phrases.
List Markets — Search public prediction-market metadata (ticker, title, phrases, rules) by free-text query and venue.
Get Market Rules — Return the rules text, phrase list and status for one market ticker.
Subscribe Phrases — Subscribe the user's account to phrases on an event (optionally delivered to a webhook). Requires an API token.
Subscribe Market — Subscribe the user's account to one mention market by ticker (CONTRACTS §18): the feed takes the market's event and compiled phrases from the market row and keeps them current when the market is re-published, so every hit is a phrase of the market on the event that settles it. Optional webhook (Standard plan+). Requires an API token.
`context` (CONTRACTS §20) opts in to context about the market, delivered alongside but
never counted as facts: "on_air" (the market's speaker is on air on any live
Market Brief — Everything relevant to one mention market in one object, every block labelled with what it is (CONTRACTS §18): the market row with the venue's latest book, the event that settles it with its sources (kind, rights, status, expected latency), the facts (phrase hits on the market's own event, newest first, with receipt hashes), the evidence-pack URL, history (base rate and last settled outcomes for the same speaker and phrase, plus the speaker profile; labelled "context, not a prediction"), related
List Market Subscriptions — The account's market subscriptions (CONTRACTS §18): each with `market_ticker`, the market's title, the live `event_id` and `phrases` it is matched on, and the webhook URL if any. Requires an API token.
Use this to answer "what markets am I following?" before subscribing again; pair with
market_brief for the state of each. Event-and-phrase subscriptions (subscribe_phrases)
are not listed here.
Get your MCP into directories
A working endpoint is step one. Directory coverage is the coordinated launch across ChatGPT, Claude, Cursor, the MCP Registry, and community indexes.
Langbot is an MCP server at mcp.langbot.ai. MCPBundles discovered 26 tools on the live endpoint.
How do I connect Langbot to my AI agent?
Use the MCP endpoint listed on this page in your MCP client configuration. One-click install pills support Claude, Cursor, VS Code, and other hosts. Copy the remote MCP URL if your client needs a manual entry.
How many tools does Langbot provide?
Operate this server?
Request removal if you do not want Langbot listed here, or claim the listing if you want to manage it on MCPBundles.
Operate Langbot?
This server appears in the MCPBundles directory. Verify you operate it to take over the listing — name, description, logo, contact email, and skill content. We email a 6-digit code to a maintainer address your server publishes in /.well-known/security.txt or /.well-known/mcpbundles.json. Free, takes about a minute.
MCPBundles publishes this directory for MCP discovery. Except where we host or operate an offering, third-party services run under their own terms. Product and company names on this page are used in a descriptive, identifying way (including under nominative fair use where applicable); they remain the property of their owners. Nothing here grants you rights in those marks, and nothing here is an offer to sell a third party's services. Terms · Request removal
More in financial
Other MCP servers in this category from the directory index
— Return the archived transcript events for an event (optionally only those after `since_seq`).
Rights note (CONTRACTS §10): for `restricted`/`unknown` sources (issuer earnings calls,
broadcaster feeds) the feed returns phrase facts only: `phrase_detected` events carry the
matched phrase, speaker label, confidence, clocks and market refs, with no `transcript_delta`
and no text beyond the phrase, so `text` will be empty. `public_domain` (federal proceedings,
Fed pressers) and `licensed` sources re
Get Recent Events — Return the most recent phrase detections (and lifecycle events) for one event or for everything live, filtered by phrases.
Get Latency Stats — Per-source latency stats (p50/p90 lag vs the public stream, in ms) from the public latency board.
Get Calendar — What's coming up: events in the next `days` days with their sources, markets, speaker keys and rights.
Use this first when asked "what's on this week", "when does Powell speak next" or to find the
`event_id` / `speaker_keys` / tickers for a later get_market_context or get_base_rates call.
`rights` tells you what you can get later: `public_domain` / `licensed` events return
transcripts; `restricted` / `unknown` return phrase facts only (CONTRACTS §10).
For events that are live right now, list_li
Get Market Context — Everything about one mention market in one call: the market row and rules, the settlement source, the current order book, the base rate for its speaker/phrase, the event it belongs to and recent mentions.
Use this as the first call for any "should I / how is X priced" question about a ticker: it
tells you what's priced (book), what the venue will settle on (rules, settlement source),
how often this speaker has said this phrase before (base_rates) and what has been heard
recently (mentions). Com
Get Base Rates — How often a speaker/phrase mention market has settled YES over `window_days` (venue settlement facts).
Use this to price a mention market or sanity-check a thesis ("does Powell ever say
'transitory'?"): `yes_rate` is the historical P(YES), `markets` the sample size,
`avg_open_price` what the market usually opened at and `mispricing` = avg(result - open
price), so a positive number means the market tends to open cheap. `by_month` shows drift.
Pass `speaker` alone for a speaker-wide rate, `phrase
Search Outcomes — The individual settled mention markets behind a base rate: ticker, speaker, phrase, result (yes/no/void), open/close prices, volume and timestamps.
Use this when you need the receipts rather than the summary: which markets settled YES, at
what opening price, how the market priced it (`yes_price_open` vs `result`), or to build
your own rate over a custom filter. `since` / `until` are ms epoch (or ISO dates if the feed
accepts them). Page with `cursor` = the previous response's `next_cursor`. Thi
Search Mentions — Langbot's own record of what has been said: `phrase_detected` facts (phrase, speaker, event_id, t_spoken, lag_vs_public_ms, market_refs) across all archived events.
Use this for "when did Jensen last say 'Blackwell'", "how early did Langbot hear it" or to
line detections up against market moves. Rows never include transcript text, only the
matched phrase (CONTRACTS §10). For the live stream of a single event use get_recent_events;
for how often a phrase *settled* YES use get_base_rates (venue f
List Speakers — Canonical speaker keys ("powell", "jensen-huang") with display name, role, org, aliases and market/mention counts.
Use this to resolve a person to the `speaker` key that get_base_rates, search_outcomes,
search_mentions and draft_strategy expect; `q` matches names and aliases. Also a quick way
to see who Langbot has history for.
Get Receipt — Look up one event's receipt: the event with its `hash`, `prev_hash`, `raw_hash` and `chain_seq`, the nearest signed checkpoint, the feed's own `verified` flag, a local recompute of the hash and the checkpoint signature check.
Use this when the user asks "prove Langbot heard that", "is this event genuine" or wants the
hash to cite. `local.ok` is this server recomputing the hash from the event (no trust in the
feed needed); `checkpoint_signature_ok` is the Ed25519 check against the published key.
Verify Receipts — Verify a batch of events (as received from the stream, a transcript or a replay): recompute every `hash` locally, walk the per-source chain, and ask the feed to confirm each one.
Use this after collecting events to prove they are unaltered. Each result has `ok` (local
recompute), `computed_hash`, `stored_hash`, and from the feed `status` (`match` /
`mismatch` / `unknown`), `chain_ok`, `chain_seq` and `checkpoint`. `chain` summarises the
local chain walk (breaks, mismatches, unhashed). Public, u
Export Replay — Pull an event's replay bundle (manifest with the signing key, every event with its receipt, signed checkpoints, optional order books) and verify it locally.
Use this to audit what was said on an event or hand the record to someone else: the
`verification` block is this server checking every hash, chain link and checkpoint
signature offline. Only the first `max_events` events are returned inline (`truncated`
tells you if there were more; raise `max_events` or filter with `since_seq` / `source_id
Sim Events — List the past events Langbot Sim can replay: each has Langbot's feed record and recorded Kalshi order books, with coverage (book rows, first/last snapshot, tickers, how many markets have settled) and whether it is a Free-plan sample.
Use this before sim_run to pick an event, or when the user asks what can be backtested.
With a token the result also carries the caller's Sim allowance (runs used and left this
period). Public.
Sim Run — Replay a past event through one of Langbot Sim's built-in strategies against the recorded Kalshi books and return the scorecard: gross/fees/net P&L, per-market results, orders, fills, fill rate, max drawdown, exposure, a latency sensitivity table (+0 / +2 s / +5 s by default) and the model's caveats.
Use this to show how a reference strategy would have fared on a real event and how much
latency costs. Strategies (by name, with `params`): `hold_none` (baseline),
`buy_yes_on_first_counting_hit` (
Attest — Record a signed attestation that the user acted on an event (`evt_id`): the feed stores `sha256(canonical(action))` bound to the event's hash and signs it.
Use this right after the user's strategy or agent takes an action in response to a
detection, so they can later prove what they did and when against Langbot's record.
`action` is the user's own JSON (<= 4 KiB): opaque to Langbot, visible only to its owner,
never a signal. Keys that look like credentials (`api_key`, `token`, `secret`, ...) ar
List Attestations — List the user's own attestations (what they recorded acting on, and when), filtered by live event, by one detection, or since a ms-epoch time.
Use this to reconstruct the user's side of the record for an event or to check whether an
action was already attested. Owner only; requires a token.
Get Receipts Key — The feed's receipt signing key: `key_id`, `algorithm` (ed25519), base64 `public_key`, plus `previous` keys after a rotation.
Use this to verify checkpoint and attestation signatures yourself, or to pin the key when
handing a replay bundle to an auditor. Public.
Get Pricing — Langbot's machine-readable pricing: plans with monthly USD price, delay on the record, limits and features, plus `per_call` prices and the `x402` section (network, asset, pay-to) for endpoints sold by the call.
Use this before recommending an upgrade, when the user asks what Langbot costs or what a
plan unlocks, or when a tool answers 402 / `payment_required`. Every tool here also carries
`_meta` with `langbot/plan_required`, `langbot/price_usd` (0 on a plan that unlocks it) and,
for per-call t
Showing the first 25 tools — full list and schemas: Directory API.
MCPBundles probed 26 tools on the live server. The tool list on this page reflects what was discovered at the last refresh — connect your client to see the full set available to your session.
What authentication does Langbot require?
Langbot may require signing in to the provider before tools can run. Connect through MCPBundles or your MCP client and complete any provider login when prompted.
Don't want this listed?
Request removal from the MCPBundles directory. We stop publishing this listing on mcpbundles.com.
MCPBundles is an independent platform built on the open Model Context Protocol standard. MCPBundles is not affiliated with, sponsored by, or endorsed by Anthropic PBC.