llms.txt
# CoreSpeed
> CoreSpeed connects AI agents to software, durable memory, usage controls, and billing through one authenticated MCP surface.
To set up an agent, give it this line: `set up https://corespeed.io/SKILL.md (read it raw, e.g. with curl)`. The skill it fetches is [SKILL.md](https://corespeed.io/SKILL.md).
## When to use CoreSpeed
Reach for CoreSpeed when an agent needs any of these, and you would rather not build them into the agent itself:
- Act inside the apps a person has already authorized — Gmail, Slack, GitHub, Notion, Linear, HubSpot and more — without registering an API application or handling a key. The agent calls the connector's tools; CoreSpeed holds the OAuth grant.
- Remember across sessions and across agents: memory is stored on the account, so a second agent (or the same one tomorrow) recalls what the first one saved.
- Generate or analyze media, search and scrape the web, or research social platforms, through built-in tools with a fixed rate card.
- Run unattended or for a team with a spend cap and a full activity log, so what it did and what it cost are on record.
How to call it: connect an MCP client to `https://api.corespeed.io/mcp` (streamable HTTP; browser sign-in, or an API key for headless use) and the connected apps, memory and built-in tools appear as namespaced tools in `tools/list`. Details: [MCP server](/docs/mcp), [Authentication](/docs/authentication), [API reference](/docs/reference).
## Product pages, as Markdown
Every page below also answers `Accept: text/markdown` at its normal URL.
- [Product overview](/index.md)
- [Pricing — plans and usage rates](/pricing.md)
- [For Sales & Revenue](/for-sales.md): Research prospects, draft outreach, and log every touch to the CRM.
- [For Marketing & Growth](/for-marketing.md): Ship campaigns, schedule social, and pull the funnel without copy-paste.
- [For Customer Support](/for-support.md): Triage tickets, draft context-aware replies, and flag churn risk early.
- [For Engineering & DevOps](/for-engineering.md): Triage the backlog, review PRs, and close out sprints from chat.
- [For Finance & Operations](/for-finance.md): Categorize expenses, reconcile payouts, and close the month on autopilot.
- [For Founders & Creators](/for-founders.md): Clear the inbox, book the calls that matter, and turn ideas into posts.
- [Blog index](/blog.md)
- [About](/about) · [Contact](/contact)
## Documentation
- [Getting started](/docs): One MCP endpoint that carries your app connections, memory, media, and spending controls to every agent you run. Connect your agent once and it gets all of it.
- [Authentication](/docs/authentication): Who is calling — a signed-in member, a member's API key, or an agent — and how each credential reaches CoreSpeed.
- [MCP server](/docs/mcp): Connect any MCP client to CoreSpeed — one remote server with browser sign-in that carries your connected apps, memory, media, web, social, and drive tools, with spend caps and an audit trail attached.
- CLI
- [CLI](/docs/cli): The cs command-line tool — sign in through the browser, mint API keys, write agent configuration, administer connectors and remote MCP servers, tie this computer, and call tools from a terminal.
- [cs login](/docs/cli/login): Sign in to CoreSpeed through the browser and store the member session every other cs command runs under.
- [cs logout](/docs/cli/logout): Remove the session cs login stored on this machine.
- [cs whoami](/docs/cli/whoami): Show the signed-in user, the organization every cs command acts on, and what you may do there.
- [cs switch](/docs/cli/switch): Change the organization every cs command acts on, without signing in again.
- [cs token](/docs/cli/token): Print a fresh session JWT to stdout for scripts and one-off requests.
- [cs keys](/docs/cli/keys): Create, list, rotate, and revoke the sk-cs- API keys that act as you in agent configuration and CI.
- [cs mcp-config](/docs/cli/mcp-config): Print an mcpServers block for the CoreSpeed MCP endpoint that carries your session token.
- [cs connectors](/docs/cli/connectors): List connectors with your connection status, open the dashboard to connect one, and manage the accounts you have connected.
- [cs remote](/docs/cli/remote): Register your organization's own MCP servers as remote connectors and administer their registrations.
- [cs mcp](/docs/cli/mcp): Find a CoreSpeed tool for a task, read its schema, and call it from a terminal as yourself — the same tools, under the same session, that an agent gets.
- [cs tie](/docs/cli/tie): Tie this computer to your organization so your agents can use its shell, files and local MCP servers, and switch what it shares.
- [cs usage](/docs/cli/usage): Show the organization's available credit and plan from the terminal.
- Connectors
- [Connectors](/docs/connectors): Connect the apps your organization already uses — by OAuth, a pasted API key, or your own OAuth app — and let every agent act through it without holding the provider credential.
- [Remote MCP servers](/docs/connectors/remote-mcp): Register your organization's own MCP servers and serve their tools through the same authenticated surface.
- **Capabilities**
- [Memory](/docs/memory): Keep durable, searchable context attached to the right member and organization across agent clients.
- [Media](/docs/media): Generate images, video, and audio from a catalog of hosted models, and ask questions about media you supply — outputs arrive as signed artifact URLs.
- [Web](/docs/web): Search the live web for ranked sources, turn one HTTPS page into clean text, and read Google Search, Google Maps, Airbnb, TripAdvisor, Amazon, and Zillow, with no provider account to connect.
- [Social](/docs/social): Read public data from TikTok, Douyin, Xiaohongshu, Instagram, YouTube, X, Reddit, Facebook, and Threads through typed, read-only tools — no platform account to connect.
- [Drive](/docs/drive): A file system for your agents — every file a tool takes or produces lives at a path in your organization's Drive, private to you or shared, kept until deleted or expiring from tmp/, mountable as an S3 bucket.
- [Capability controls](/docs/capability-controls): Narrow which built-in capabilities appear in your own tool list, and understand the organization-level ceiling above it.
- **Operate**
- [Activity & audit](/docs/activity): Inspect the organization-scoped trail for MCP tools and control-plane actions.
- [Billing & credits](/docs/billing): Credits, the organization wallet, API-key spend caps, and the one billing boundary applied to every metered CoreSpeed tool.
- [Approvals](/docs/approvals): Write in plain English what needs your sign-off. An action you said to check pauses, asks you, and runs only once you say yes.
- [Error handling](/docs/errors): Decide what to retry, what to fix, and when to send a user back to the dashboard after a CoreSpeed failure.
## API Reference
- [Overview](/docs/reference): The authenticated CoreSpeed HTTP surface — MCP JSON-RPC methods, connector discovery, and the error contract.
- [initialize](/docs/reference/mcp/initialize): Negotiate the MCP protocol version and read the CoreSpeed server capabilities.
- [tools/list](/docs/reference/mcp/tools-list): Enumerate every tool visible to the authenticated caller on the CoreSpeed MCP endpoint.
- [tools/call](/docs/reference/mcp/tools-call): Invoke one namespaced CoreSpeed tool and read its result or isError payload.
- [List connectors](/docs/reference/connectors/list): Every connector visible to the caller, with connection status and account state.
- [Get connector](/docs/reference/connectors/get): One connector entry; 404 covers both an unknown id and a connector this environment does not serve.
- [Disconnect account](/docs/reference/connectors/delete): Disconnect your own private account on an OAuth connector. Remote MCP deregistration lives on its own path.
- [manage__* tools](/docs/reference/tools/manage): The account-management tools on the MCP surface — API keys, agents, connected accounts, environments, session, remote MCP servers, approval waits, and feedback — with arguments and who may call each.
- [Errors & status codes](/docs/reference/errors): The single source for CoreSpeed HTTP status codes, error codes, error envelopes, and MCP isError results.
## Blog
- [Memory that keeps its receipts](/blog/memory-that-keeps-its-receipts): Extracted facts answered 84.4% of LongMemEval-S. Linking each fact back to the conversation it came from raised that to 89.6%, for 20% more storage.
- [TinyFish, now on CoreSpeed](/blog/tinyfish-on-corespeed): Web agents, search, saved sign-ins and monitors from TinyFish, in every agent you use, through one CoreSpeed connection.
- [CoreSpeed for SEO](/blog/corespeed-for-seo): Keyword, rank and audit data for any agent, billed per call in credits.
- [Xbrain: Give Your Agent a Sense of Your Taste](/blog/introducing-xbrain): Turn your X bookmarks into searchable design context for your agent.
- [A meeting brief from your mail, calendar and notes, in one request](/blog/a-meeting-brief-from-mail-calendar-and-notes): How an agent builds a meeting brief from Gmail, Google Calendar and Notion through private connections, remembers your format, and never holds a token.
- [A support agent that issues refunds, with a person on the big ones](/blog/support-refunds-with-an-approval-step): A worked example. A support agent reads the ticket, checks the charge in Stripe and issues the refund. Refunds over $100 pause for a person in Slack.
- [Built-in web search vs wiring a search API into your agent](/blog/built-in-web-search-vs-connecting-a-search-api): CoreSpeed's web tools return ranked sources and clean page text with no provider account, billed per request. A direct provider key gives more knobs.
- [Competitive research: search for sources, then read what matters](/blog/competitive-research-with-search-and-scrape): A playbook for an agent that finds competitor sources with web__search, reads the few pages that matter with web__scrape, and writes the brief to Notion.
- [Connect Claude Code to Notion, Slack and GitHub with one MCP server](/blog/connect-claude-code-to-your-apps-with-one-mcp-server): Add the CoreSpeed MCP server to Claude Code at user scope, sign in once, connect Notion, Slack and GitHub in the dashboard, and verify with manage__whoami.
- [Continue a task in a different agent without re-explaining](/blog/continue-a-task-in-a-different-agent): What carries over when you move a task from Claude Code to Codex through CoreSpeed, memory, connections and policy, and what does not, the conversation itself.
- [CoreSpeed FAQ: plans, credits, clients, security, what it does not do](/blog/corespeed-faq): Short answers on what CoreSpeed is, which agents work, what is free, where tokens live, who approves, whether it runs agents, and how to leave.
- [Exactly once: how an approved agent action runs with frozen arguments](/blog/exactly-once-approvals): A held call is frozen, the agent waits on one id, and approval hands the call to that waiting request once. Re-sending the call makes a second card.
- [Generating launch images and short videos from an agent](/blog/generate-launch-assets-with-media-tools): How an agent picks a hosted model, generates a launch image, chains it into a short video through its Drive path, and collects signed URLs. Billed only after success.
- [How memory search works: hybrid retrieval, excerpts and result budgets](/blog/how-memory-search-works): memory__search_memory fuses exact and relaxed lexical matching, entity aliases and vector similarity into one ranking, then bounds each row and the result.
- [How to add one MCP server to Cursor for apps, memory and web search](/blog/add-corespeed-to-cursor): Add CoreSpeed to ~/.cursor/mcp.json, click Needs login, and Cursor gets your connected apps, durable memory, live web search and media tools from one server.
- [How to connect ChatGPT to your apps with a custom connector](/blog/use-corespeed-in-chatgpt-developer-mode): Turn on Developer mode in ChatGPT (Plus and Pro on the web), create a connector for api.corespeed.io/mcp with OAuth, and sign in to reach your apps and memory.
- [How to connect Codex to your apps and give it memory](/blog/connect-codex-to-your-apps-and-memory): Register CoreSpeed in Codex with codex mcp add, sign in, connect your apps, and add memory rules to AGENTS.md so Codex recalls decisions across sessions.
- [How to connect two accounts of the same app to one agent](/blog/connect-two-accounts-of-the-same-app): Connect two Slack workspaces or two X handles under one connector, give each an alias, and pass the account argument so each call lands in the right one.
- [How to give your agent memory that survives sessions and clients](/blog/give-your-agent-memory-across-sessions): Two standing rules, a scope choice, and a habit of saving decisions give any agent memory that outlives the session and the client it ran in.
- [How to register your own MCP server for your whole organization](/blog/register-your-own-mcp-server-as-an-org-connector): An org admin registers any HTTPS MCP server as a remote connector; its tools join tools/list under the org__ prefix with the same sign-in, caps and trail.
- [How to run a headless agent or CI job with a CoreSpeed API key](/blog/api-keys-for-headless-agents-and-ci): Choose a member key (sk-cs-) or an agent key (sk-csa-), read it from CORESPEED_API_KEY, cap its monthly spend, and verify the key with memory__list_memory.
- [How to set a monthly budget for an agent](/blog/set-a-monthly-budget-for-an-agent): Give the agent its own key with a monthly spend cap in credits, know what the refusal looks like, and read what each run cost in Activity.
- [How to share one app connection with your whole team](/blog/share-an-app-connection-with-your-team): Connect an app once as a shared account so every teammate's agent can act through it. Who can share, who can remove, and when to keep a connection private.
- [How to use your apps and memory inside claude.ai and Claude Desktop](/blog/use-corespeed-in-claude-ai-and-claude-desktop): Add CoreSpeed as a custom connector on the Pro, Max, Team or Enterprise plan. It follows your Claude account to desktop and mobile with your apps and memory.
- [How to write your first approval policy (with examples)](/blog/write-your-first-approval-policy): Write three kinds of sentence (fine, ask, never), pick a default behavior, run Test policy, connect Slack and turn the gate on. Reads are never gated.
- [HTTP 200 with isError true: the most common MCP integration bug](/blog/http-200-with-iserror-true): A CoreSpeed tool call can fail inside an HTTP 200. How the two error layers differ, which codes to retry, and which ones need a person instead.
- [Linking memories: a small graph that keeps decisions connected](/blog/a-memory-graph-with-links): Memory links are directed edges with a kind and optional weight. They are idempotent to write, visible only when both ends are, and capped at 1,000 per memory.
- [Market scans with listing tools: Zillow, Airbnb, TripAdvisor and Maps](/blog/real-estate-and-travel-scans-with-listing-tools): Read Zillow, Airbnb, TripAdvisor and Google Maps as structured records. Each call collects live in 10 to 60 seconds and bills only the rows it gathered.
- [Member key vs agent key vs browser sign-in: which to use](/blog/member-key-vs-agent-key-vs-browser-sign-in): Four credentials, two kinds of principal. What each one reaches, where spend caps attach, and which credential an unattended agent should carry.
- [Memory isolation with row-level security, not application code](/blog/memory-isolation-with-row-level-security): Caller identity comes from auth, each memory carries its scope, and Postgres row-level security enforces the boundary. Nothing in the request body can cross it.
- [needs_reauth explained: what it means and what to do](/blog/needs-reauth-explained): needs_reauth means an account's upstream grant can no longer be used. Tools stay visible, calls fail, and the fix is reauthorizing in the dashboard.
- [OAuth connector vs pasting an API key: which to use for each app](/blog/oauth-connector-vs-pasting-an-api-key): Most CoreSpeed connectors use OAuth. Some take a pasted provider key, a few take only that. How the index tells them apart, and how rotation and reauth differ.
- [One MCP server vs a config file per agent](/blog/one-mcp-server-vs-a-config-file-per-agent): When a per-client list of local MCP servers is enough, where it stops scaling, and what changes once a team points every client at one remote server.
- [One org, many agents: Claude Code, Codex and a bot on one rulebook](/blog/one-org-many-agents): How a team keeps each member's own client, shares app connections, runs one approval policy and one ledger, and gives bots an identity of their own.
- [payment_required and key_spend_limit_exceeded: what still works](/blog/payment-required-and-spend-limit-errors): payment_required is an organization wallet hold; key_spend_limit_exceeded is one key's monthly cap. Both arrive as HTTP 200 with isError true.
- [Pull requests for shared memory: proposals, archive and versions](/blog/pull-requests-for-shared-memory): A correction next to a stale shared memory does not stop the old one being retrieved. Archive it with a successor, or propose a fix for its author to merge.
- [Running an unattended job you operate with a CoreSpeed agent key](/blog/an-unattended-job-with-an-agent-key): Give a cron job or server process you run its own agent identity and sk-csa key: shared connections only, a monthly cap, approvals routed to its owner.
- [Shared memory vs CLAUDE.md and AGENTS.md files](/blog/shared-memory-vs-instruction-files): Instruction files hold the rules an agent follows, per repository or per machine and per client. Org-owned memory holds searchable facts that follow you across both. Use both.
- [Smart Approval vs client-side permission prompts](/blog/smart-approval-vs-client-side-permission-prompts): Claude Code's auto mode and Codex's Approve for me guard one client. Smart Approval gates the tools for every client. When each one is the right choice.
- [Social listening with read-only tools and no platform accounts](/blog/social-listening-with-read-only-tools): How an agent tracks mentions on X, Reddit, YouTube, TikTok and Xiaohongshu through read-only tools, with no platform login, and saves findings to shared memory.
- [The cs CLI in ten minutes: login, keys, tools, usage](/blog/the-cs-cli-in-ten-minutes): Install @corespeed/cs on Node 20 or Bun, sign in with cs login, mint and rotate API keys, print MCP config, search and call tools, check usage, and read exit codes in scripts.
- [The hardest problem in multi-agent teams is scoping](/blog/the-hardest-problem-in-multi-agent-teams-is-scoping): Past one person, the hard question is scope: who may touch what, on whose behalf, with what budget, with what record. The organization has to answer it.
- [What are spend caps for AI agents, and how do they work?](/blog/what-are-spend-caps-for-ai-agents): Spend caps stop an agent loop from draining a budget. On CoreSpeed an org wallet threshold and per-key monthly caps refuse metered calls before a tool runs.
- [What does human-in-the-loop mean for AI agents in 2026?](/blog/what-is-human-in-the-loop-for-ai-agents): Human-in-the-loop means a person decides named actions before they run, under a policy written in advance. Reads never wait, and an approved call runs once.
- [What is a plain-English approval policy?](/blog/what-is-a-plain-english-approval-policy): Sentences, read in full for every write an agent makes. Every save is a version, Test policy runs a hypothetical call, and a default you pick covers the gaps.
- [What is a remote MCP server, and how do you connect one?](/blog/what-is-a-remote-mcp-server): A remote MCP server is reached over HTTPS instead of run as a local process. How sign-in works, and how
