Skip to main content

Connect agents with MCP

MCP (Model Context Protocol) is the open standard AI agents use to call external tools. Corte ships an MCP server, so an agent like Claude Code can generate media, run your Applications, and execute your Workflows directly — no glue code, no SDK. You describe the outcome; the agent picks the tools.

Connect from Claude Code​

The server speaks Streamable HTTP at https://api.corte.so/mcp:

claude mcp add --transport http corte https://api.corte.so/mcp

A browser window opens, you sign in to Corte, and a screen lists exactly what the agent will be able to reach. Approve it and the connection is live — no key to copy, nothing to paste. Claude Code, claude.ai and Cursor all take this route; any client that follows the MCP authorization spec does.

Every app you approve appears in Settings → API alongside your keys, with what it can reach and when it last acted. Disconnect one there and it stops working immediately.

Without a browser​

Scripts, CI jobs and servers authenticate with an API key instead:

claude mcp add --transport http corte https://api.corte.so/mcp \
--header "Authorization: Bearer corte_sk_…"

:::warning Agents spend real credits MCP tools run as your account — every generation or workflow run an agent starts costs the same credits as clicking Generate yourself. Keep an eye on Settings → Usage, and give agents their own key so you can revoke access independently. :::

An agent sees only the tools its grant covers, using the same scopes a key does. A key limited to content:rw lists the content tools and nothing else — it cannot reach generate or run_workflow, so it cannot spend credits at all. That is the tightest way to hand an agent a CMS without handing it your balance.

Available tools​

Every tool is annotated: reads are marked read-only, deletes destructive, so a host that honours the annotations (Claude does) lets an agent read without asking each time and puts writes behind approval. The table runs in the same order as the reference page: account and credits, projects, media, brand, workflows, agents, then the Growth queue.

ToolWhat it does
get_creditsYour remaining credit balance — lets an agent check before spending.
list_projectsThe projects you can work in and what each can spend — so "use the Playground project" resolves to the right projectId.
get_accountYour plan, credits, storage and whether the account can spend — the answer to "what plan am I on" and "why was that refused".
get_usageWhat you spent, by kind and by day, plus the activity feed — "what did that charge buy".
list_plansThe plan ladder with prices, monthly credits and storage, current plan marked — so an agent can say what an upgrade costs.
create_topup_checkoutA payment link for a top-up, for the agent to hand you when the balance runs dry.
create_subscription_checkoutA payment link for a plan.
create_billing_portal_linkA link to the billing portal, for the agent to hand you: change the card, download invoices, switch to yearly, cancel.
get_projectOne project in full — who is in it, who pays for it, what it can spend.
create_projectA new project of your own, ready for the other tools' projectId.
update_projectRename a project, change its description or picture.
invite_to_projectInvite someone by email — sends the invite, they join as a contributor when they accept.
cancel_project_inviteWithdraw an invite nobody has accepted yet.
remove_project_memberTake a member out of a project; their work stays.
get_project_usageThe project's spend, entry by entry and totalled per member.
list_modelsThe model catalog with capabilities and per-unit pricing.
generateStart a generation with any model (image, video, audio, upscale).
get_generationPoll a generation for status and result URLs.
transcribe_assetSubtitles from a stored audio or video file: text, timed segments, SRT and VTT.
get_transcriptionPoll a transcription for its status and subtitles.
upload_assetImport media from a URL; returns a storageId other tools accept.
list_applicationsThe 46 curated Applications and their input requirements.
run_applicationRun an Application with images/video/text and an optional preset.
get_brandThe project's brand — palette, type, photography direction, voice — and its DESIGN.md.
get_brand_design_mdThe brand as DESIGN.md text, ready for a repo or a system prompt.
update_brandReplace the brand doc in full.
import_brandPull a brand off a website, Figma file, HTML, screenshots or a DESIGN.md, staged for review.
get_brand_importPoll an import for its status and the field-by-field diff.
list_brand_importsThe project's imports and what was applied from each.
apply_brand_importAccept an import — every change, or just the fields the person chose.
create_brand_directionsThree candidate brand directions from a brief, each with pictures and first copy.
get_brand_directionsPoll a brand-creation session for its directions.
accept_brand_directionMake a direction the project's brand.
list_workflow_node_typesEvery node type with its ports and settings — the vocabulary for composing a pipeline.
create_workflowSave a new pipeline from a graph, checked before it is stored.
list_workflowsYour saved workflow pipelines.
get_workflowOne workflow's graph — including its input node ids.
run_workflowExecute a workflow with custom inputs.
get_workflow_runPoll a run for status and per-node results.
update_workflowRename a workflow or replace its graph.
delete_workflowRemove a workflow and its run history.
list_workflow_runsA workflow's run history, newest first.
cancel_workflow_runStop a run that is still going.
list_agentsThe automations you have set up, with their schedule and when each last ran.
get_agentOne automation in full — its prompt, schedule, scopes and last run.
create_agentSet up a standing automation that runs a prompt on a schedule, on a content event, or on demand.
run_agentStart an automation now, without waiting for its schedule.
list_agent_runsWhat an automation has been doing — every run, its status, what it made and what it cost.
get_agent_runOne run in full, with its report and transcript.
reply_to_agent_runAnswer a run that ended on a question, or ask for a change to what it produced.
cancel_agent_runStop a run that is still going.
list_agent_outputsEverything an automation has made — the documents and posts, newest first.
list_agent_memories · add_agent_memory · delete_agent_memoryWhat an automation has learned and been told, and the way to correct it without rewriting its prompt.
delete_agentRemove an automation and its run history for good.
list_growth_channelsThe social accounts a project can publish to.
create_growth_postDraft a social post into your review lane — you approve it in Growth.
list_growth_postsThe publish queue across every state, with click counts.
get_growth_postOne post: text, media, state, live URL, last error.
update_growth_postFix a draft or a post awaiting review.
decline_growth_postSay no to a draft; it stays as history.
unschedule_growth_postPull a scheduled post back before it goes out.
retry_growth_postSend a failed post again once the cause is fixed.
delete_growth_postRemove a post from the queue.
get_growth_post_analyticsClicks on a post's links and the platform's engagement numbers.
list_growth_campaignsCampaigns — one story's posts, grouped.
create_growth_campaignStart a campaign to draft posts into.
get_growth_campaignA campaign's posts and combined clicks.
update_growth_campaignRename a campaign.
delete_growth_campaignDelete a campaign; its posts stay.
list_catalog_productsThe products in the project's ad catalog — price, photos, and whether each is ready for ads.
sync_catalogBring the connected store's products into the ad catalog; updates in place.
prepare_catalog_productReady a product's photos for ads: cut-outs and a reading of each photo (credits per photo).
list_ad_layoutsThe ad layouts and house styles, and what each layout needs from a product.
create_ad_setStart an ad set of product ads in one look, with its public catalog feed URL.
list_ad_setsYour ad sets, their counts, feed URL, and what a refresh would change.
generate_catalog_adsMake one product's ads in an ad set — layouts, art-directed designs, or both.
list_adsAn ad set's ads, by product or status, with any issues the checker found.
update_adApprove, reject or replace an ad's document; approving publishes it to the feed.
redesign_adRework one ad from a plain instruction; returns what changed.
refresh_ad_setBring an ad set up to date after the look, photos or offer changed — free; edited ads stay.
rewrite_product_copyA new headline and copy for one product's ads; edited ads stay.

Full parameter and return shapes for each tool are in the MCP tools reference.

The typical agent flow​

Ask an agent for something like "run my Product Reshoot workflow on this photo" and a well-behaved run looks like:

  1. upload_asset { url } → a committed storageId.
  2. get_workflow { workflowId } → discover the graph's input node ids.
  3. run_workflow { workflowId, inputs: { productPhoto: { storageId }, concept: { text: "…" } } }
  4. Poll get_workflow_run — outputs land in run.nodeResults[*].output.value as media URLs and text.

The same id-keyed input convention as the REST endpoint applies — see Workflows via the API for the details.

When something fails​

Tool errors carry the same codes as the REST API (overview). Where the account is what stopped the call — the balance ran dry, the address is unconfirmed, the model needs a purchase, storage is full — the agent gets the code, the balance and a sentence saying what to do, so it can tell you rather than stall. For a dry balance that means a top-up link it hands you; you pay in your browser, and the credits land within a minute. Details in the MCP tools reference. If generations themselves fail, see Generation issues.