--- title: "Claude API (Anthropic) MCP server for AI agents - actionlake" description: "Give Claude 113 Claude API (Anthropic) actions through one MCP server: deployments. Your agent never holds the credential. 100,000 calls a month free." canonical: https://actionlake.com/connectors/anthropic-claude-api updated: 2026-10-10 --- Connectors · AI # Claude API (Anthropic) MCP server for AI agents. actionlake is an MCP gateway: one remote MCP server that connects Claude and other MCP clients to Claude API (Anthropic) and the rest of your systems. It gives your agent 113 Claude API (Anthropic) actions (52 reads, 46 writes and 15 that wait for a confirmation) and holds the Claude API (Anthropic) credential itself, so the agent never does. It paces Claude API (Anthropic)'s limits and records every call. [Connect Claude API (Anthropic) free](https://actionlake.com/console?start=sandbox) [See pricing](https://actionlake.com/pricing) 100,000 free calls a month. No card, no email, no password: press Start a free sandbox on the next screen and keep the key it shows you. `claude mcp add --transport http actionlake https://connect.actionlake.com/mcp` the first call ``` do { action: "anthropic.files.list", intent: "check which Claude API (Anthropic) account is connected" } # List Files · read · 1 upstream call ``` - 113 actions in one signed file - 52 read look, never change - 46 write create and update - 15 wait for a yes destructive, confirmed first Set up ## Connect Claude API (Anthropic) in three steps. 1. 01 ### Add actionlake to Claude. Run the command once in Claude Code; its consent page can create a free sandbox: 100,000 calls a month, no card. In the Claude app, add a custom connector with the URL below and paste your key on the consent page. Creating a new sandbox from the Claude app is not yet tested by us. `https://connect.actionlake.com/mcp` 2. 02 ### Connect Claude API (Anthropic). API key. No consent screen: your agent can connect it for you, but the key then passes through your chat. To keep it out of the model's context, paste it in the console instead. actionlake stores what you paste write-only, sealed under your tenant's key. - `api_key` Anthropic API key (sk-ant-api…), scoped to one workspace ``` account { op: "connections.requirements", service: "anthropic" } # free: exactly what to paste, and what is found for you ``` 3. 03 ### Make the first call. Check before a write. Ask in plain words; Claude picks the action. Before anything changes, `check` validates the call and prices it, for free. ``` do { action: "anthropic.files.list" } # list files · read · 1 upstream call check { action: "anthropic.files.delete", input: { … } } # free: inputs, permissions, cost and confirmations, nothing sent ``` Claude API (Anthropic) at a glance **Sign-in** API key. No consent screen: your agent can connect it for you, but the key then passes through your chat. To keep it out of the model's context, paste it in the console instead. **You provide** `api_key` **Limits we pace to** at most 16 at a time, per connection **API covered** Claude API (Anthropic) API 2023-06-01`api.anthropic.com` **Built from** Claude API (Anthropic)'s documentation, read 1 October 2026[GitHub](https://raw.githubusercontent.com/anthropics/anthropic-sdk-python/main/scripts/mock-spec.json.gz) **Service file** Version 1.0.0, signed (Ed25519) **Cached reads** 3 of 52 reads, for up to an hour; a write clears what it affects **Shared entities** 1, so ask can line Claude API (Anthropic) up with your other systems **Data** Processed and stored in Sydney, Australia. No US region yet Capabilities ## What an AI agent can do in Claude API (Anthropic). 113 actions in 27 areas. Every one carries its effect, so a key can allow the reads and hold the rest. - **52** read look, never change - **46** write create and update - **15** destructive wait for a confirmation 1. ### Deployments 8 2 read · 6 write · 0 destructive List Deployments · Create Deployment · Update Deployment · Get Deployment 2. ### Environment work 8 4 read · 4 write · 0 destructive List Work Items · Acknowledge Work · Get Work Item · Record Heartbeat 3. ### Vault credentials 7 3 read · 3 write · 1 destructive List Credentials · Create Credential · Get Credential · Update Credential 4. ### Environments 6 2 read · 3 write · 1 destructive Create Environment · List Environments · Archive Environment · Get Environment 5. ### Memory stores 6 2 read · 3 write · 1 destructive Create a memory store · Retrieve a memory store · Update a memory store · Delete a memory store 6. ### Sessions 6 2 read · 3 write · 1 destructive Create Session · List Sessions · Archive Session · Get Session 7. ### Tunnels 6 3 read · 2 write · 1 destructive List Tunnels · Create Tunnel · Get Tunnel · Archive Tunnel 8. ### Vaults 6 2 read · 3 write · 1 destructive Create Vault · List Vaults · Archive Vault · Get Vault **19 more areas · 60 actions** - Agents **5** - Dreams **5** - Memories **5** - Message batches **5** - Session resources **5** - User profiles **5** - Tunnel certificates **4** - Files **3** - Memory versions **3** - Session threads **3** - Skill versions **3** - Skills **3** - Deployment runs **2** - Messages **2** - Models **2** - Session events **2** - Agent versions **1** - Completions **1** - Session thread events **1** Top actions ## The Claude API (Anthropic) API, shaped for agents. The ids your agent types, what each does, and the one name it shares with other systems. 40 of Claude API (Anthropic)'s 40 action names are shared with at least one other connector, so an agent learns them once. | Action | What it does | Effect | Same name in | | --- | --- | --- | --- | | `anthropic.files.list` List Files | List Files | read | `content.file.list` 77 connectors | | `anthropic.files.get` Get File Metadata | Get File Metadata | read | `content.file.get` 74 connectors | | `anthropic.files.delete` Delete File | Delete File | destructive | `content.file.delete` 68 connectors | | `anthropic.message_batches.list` List Message Batches | List all Message Batches within a Workspace. | read | `automation.background_job.list` 34 connectors | | `anthropic.message_batches.create` Create a Message Batch | Send a batch of Message creation requests. | write | `automation.background_job.create` 29 connectors | | `anthropic.agents.create` Create Agent | Create a versioned agent: model, system prompt, tools, MCP servers and skills; returns the agent (bind `id`, and `version` to pin it). | write | `infra.ml_agent.create` 6 connectors | | `anthropic.models.get` Get a Model | Get a specific model. | read | `infra.ml_model.get` 13 connectors | | `anthropic.tunnels.list` List Tunnels | The Tunnels API is in research preview. | read | `infra.vpn_connection.list` 5 connectors | | `anthropic.deployments.list` List Deployments | List Deployments | read | `automation.schedule.list` 18 connectors | | `anthropic.deployments.create` Create Deployment | Create Deployment | write | `automation.schedule.create` 16 connectors | | `anthropic.tunnels.create` Create Tunnel | The Tunnels API is in research preview. | write | `infra.vpn_connection.create` 3 connectors | | `anthropic.agents.get` Get Agent | Get Agent | read | `infra.ml_agent.get` 6 connectors | Try it ## Prompts to paste into Claude. Each one names the Claude API (Anthropic) actions it will call, so you know what it reads and what it changes before you send it. 1. Read read Which Claude API (Anthropic) files changed this week, and who changed them? calls `anthropic.files.list` 2. Write write Create a Message Batch in Claude API (Anthropic). Show me exactly what you will send before you save it. calls `anthropic.message_batches.create` Governance for Claude API (Anthropic) ## Destructive actions wait for a yes. 15 Claude API (Anthropic) actions are destructive. These are all of them: - Delete File `anthropic.files.delete` - Cancel a Message Batch `anthropic.message_batches.cancel` - Delete a memory `anthropic.memories.delete` - Delete a memory store `anthropic.memory_stores.delete` - Delete a Message Batch `anthropic.message_batches.delete` - Delete Credential `anthropic.vault_credentials.delete` - Rotate Tunnel Token `anthropic.tunnels.rotate_token` - Delete Skill `anthropic.skills.delete` - Delete Vault `anthropic.vaults.delete` - Delete Session `anthropic.sessions.delete` - Delete Environment `anthropic.environments.delete` - Delete Skill Version `anthropic.skill_versions.delete` - Redact a memory version `anthropic.memory_versions.redact` - Delete Session Resource `anthropic.session_resources.remove` - Cancel a Dream `anthropic.dreams.cancel` A destructive action runs only with `confirm: true`. Under the `human` policy the agent cannot confirm for itself: the run stops until a named person approves, in the console or from a signed link in an email. The cap on a key is refused before the call is sent, never after. do anthropic.files.delete confirm_policy: **human** Delete File? **intent:** the user asked me to delete File run suspended · waiting for a named person · key _anthropic-agent_ cannot confirm itself In the console, or from an emailed signed link. a key for this connector ``` account { op: "keys.create", name: "anthropic-agent", allowed_actions: ["anthropic.*"], cap_calls: 2000, confirm_policy: "human_for_writes", approval_rules: [ { match: "anthropic.files.delete", approver: "lead@example.com" }, { match: "anthropic.message_batches.cancel", approver: "lead@example.com" }, { match: "anthropic.memories.delete", approver: "lead@example.com" } ] } ``` Illustrative. Scoped to Claude API (Anthropic) only, capped at 2,000 upstream calls, every write confirmed, and the destructive ones approved by a named person. Cost ## What Claude API (Anthropic) costs on actionlake. Every Claude API (Anthropic) action is one upstream call, the only thing actionlake bills. Searching, `check`, `cost`, `account` and any call Claude API (Anthropic) refused with 401, 403, 429 or a 5xx are free. Paid plans start at US$49 a month for 250,000 calls, with overage from US$0.30 down to US$0.18 per 1,000. - 1 upstream call per action all 113 Claude API (Anthropic) actions - 113 actions page each page of a long list is one call - 3 reads can be cached a cache hit costs nothing - 100,000 calls a month, free paid plans from US$49 a month for 250,000 [See pricing](https://actionlake.com/pricing) Action index ## Every Claude API (Anthropic) action. 113 actions by area: the id your agent calls, what it does, and its effect. **Deployments 8** - `anthropic.deployments.archive` Archive Deployment write - `anthropic.deployments.create` Create Deployment write - `anthropic.deployments.get` Get Deployment read - `anthropic.deployments.list` List Deployments read - `anthropic.deployments.pause` Pause Deployment write - `anthropic.deployments.run` Run Deployment Now write - `anthropic.deployments.unpause` Unpause Deployment write - `anthropic.deployments.update` Update Deployment write **Environment work 8** - `anthropic.environment_work.ack` Acknowledge Work write - `anthropic.environment_work.get` Get Work Item read - `anthropic.environment_work.heartbeat` Record Heartbeat write - `anthropic.environment_work.list` List Work Items read - `anthropic.environment_work.poll` Poll for Work read - `anthropic.environment_work.stats` Get Queue Statistics read - `anthropic.environment_work.stop` Stop Work write - `anthropic.environment_work.update` Update Work Item write **Vault credentials 7** - `anthropic.vault_credentials.archive` Archive Credential write - `anthropic.vault_credentials.create` Create Credential write - `anthropic.vault_credentials.delete` Delete Credential destructive - `anthropic.vault_credentials.get` Get Credential read - `anthropic.vault_credentials.list` List Credentials read - `anthropic.vault_credentials.update` Update Credential write - `anthropic.vault_credentials.validate_mcp_oauth` Validate Credential read **Environments 6** - `anthropic.environments.archive` Archive Environment write - `anthropic.environments.create` Create Environment write - `anthropic.environments.delete` Delete Environment destructive - `anthropic.environments.get` Get Environment read - `anthropic.environments.list` List Environments read - `anthropic.environments.update` Update Environment write **Memory stores 6** - `anthropic.memory_stores.archive` Archive a memory store write - `anthropic.memory_stores.create` Create a memory store write - `anthropic.memory_stores.delete` Delete a memory store destructive - `anthropic.memory_stores.get` Retrieve a memory store read - `anthropic.memory_stores.list` List memory stores read - `anthropic.memory_stores.update` Update a memory store write **Sessions 6** - `anthropic.sessions.archive` Archive Session write - `anthropic.sessions.create` Create Session write - `anthropic.sessions.delete` Delete Session destructive - `anthropic.sessions.get` Get Session read - `anthropic.sessions.list` List Sessions read - `anthropic.sessions.update` Update Session write **Tunnels 6** - `anthropic.tunnels.archive` Archive Tunnel write - `anthropic.tunnels.create` Create Tunnel write - `anthropic.tunnels.get` Get Tunnel read - `anthropic.tunnels.list` List Tunnels read - `anthropic.tunnels.reveal_token` Reveal Tunnel Token read - `anthropic.tunnels.rotate_token` Rotate Tunnel Token destructive **Vaults 6** - `anthropic.vaults.archive` Archive Vault write - `anthropic.vaults.create` Create Vault write - `anthropic.vaults.delete` Delete Vault destructive - `anthropic.vaults.get` Get Vault read - `anthropic.vaults.list` List Vaults read - `anthropic.vaults.update` Update Vault write **Agents 5** - `anthropic.agents.archive` Archive Agent write - `anthropic.agents.create` Create Agent write - `anthropic.agents.get` Get Agent read - `anthropic.agents.list` List Agents read - `anthropic.agents.update` Update Agent write **Dreams 5** - `anthropic.dreams.archive` Archive a Dream write - `anthropic.dreams.cancel` Cancel a Dream destructive - `anthropic.dreams.create` Create a Dream write - `anthropic.dreams.get` Get a Dream read - `anthropic.dreams.list` List Dreams read **Memories 5** - `anthropic.memories.create` Create a memory write - `anthropic.memories.delete` Delete a memory destructive - `anthropic.memories.get` Retrieve a memory read - `anthropic.memories.list` List memories read - `anthropic.memories.update` Update a memory write **Message batches 5** - `anthropic.message_batches.cancel` Cancel a Message Batch destructive - `anthropic.message_batches.create` Create a Message Batch write - `anthropic.message_batches.delete` Delete a Message Batch destructive - `anthropic.message_batches.get` Retrieve a Message Batch read - `anthropic.message_batches.list` List Message Batches read **Session resources 5** - `anthropic.session_resources.add` Add Session Resource write - `anthropic.session_resources.get` Get Session Resource read - `anthropic.session_resources.list` List Session Resources read - `anthropic.session_resources.remove` Delete Session Resource destructive - `anthropic.session_resources.update` Update Session Resource write **User profiles 5** - `anthropic.user_profiles.create` Create User Profile write - `anthropic.user_profiles.create_enrollment_url` Create Enrollment URL write - `anthropic.user_profiles.get` Get User Profile read - `anthropic.user_profiles.list` List User Profiles read - `anthropic.user_profiles.update` Update User Profile write **Tunnel certificates 4** - `anthropic.tunnel_certificates.archive` Archive Tunnel Certificate write - `anthropic.tunnel_certificates.create` Create Tunnel Certificate write - `anthropic.tunnel_certificates.get` Get Tunnel Certificate read - `anthropic.tunnel_certificates.list` List Tunnel Certificates read **Files 3** - `anthropic.files.delete` Delete File destructive - `anthropic.files.get` Get File Metadata read - `anthropic.files.list` List Files read **Memory versions 3** - `anthropic.memory_versions.get` Retrieve a memory version read - `anthropic.memory_versions.list` List memory versions read - `anthropic.memory_versions.redact` Redact a memory version destructive **Session threads 3** - `anthropic.session_threads.archive` Archive Session Thread write - `anthropic.session_threads.get` Get Session Thread read - `anthropic.session_threads.list` List Session Threads read **Skill versions 3** - `anthropic.skill_versions.delete` Delete Skill Version destructive - `anthropic.skill_versions.get` Get Skill Version read - `anthropic.skill_versions.list` List Skill Versions read **Skills 3** - `anthropic.skills.delete` Delete Skill destructive - `anthropic.skills.get` Get Skill read - `anthropic.skills.list` List Skills read **Deployment runs 2** - `anthropic.deployment_runs.get` Get Deployment Run read - `anthropic.deployment_runs.list` List Deployment Runs read **Messages 2** - `anthropic.messages.count_tokens` Count tokens in a Message read - `anthropic.messages.create` Create a Message write **Models 2** - `anthropic.models.get` Get a Model read - `anthropic.models.list` List Models read **Session events 2** - `anthropic.session_events.list` List Events read - `anthropic.session_events.send` Send Events write **Agent versions 1** - `anthropic.agent_versions.list` List Agent Versions read **Completions 1** - `anthropic.completions.create` Create a Text Completion write **Session thread events 1** - `anthropic.session_thread_events.list` List Session Thread Events read Questions ## The Claude API (Anthropic) MCP server, answered. **Is there a Claude API (Anthropic) MCP server for Claude?** Yes. actionlake is an MCP gateway: a remote MCP server that gives Claude, or another MCP client, 113 Claude API (Anthropic) actions across deployments, environment work, vault credentials and environments, through the same eight tools it uses for every other connector. actionlake is independent; it is not made or endorsed by Claude API (Anthropic). **What do I need to connect Claude API (Anthropic)?** Anthropic API key (sk-ant-api…), scoped to one workspace. The free call account { op: "connections.requirements", service: "anthropic" } lists exactly what to paste. **Can an AI agent delete or change things in Claude API (Anthropic)?** Only if its key allows it, and only with an explicit confirmation. 15 Claude API (Anthropic) actions are destructive, including delete File, cancel a Message Batch and delete a memory. Each runs only with confirm: true. Under the human confirm policy, or an approval rule on the key, the run waits until a named person approves it. **Will an AI agent use up my Claude API (Anthropic) API limit?** actionlake paces every call to the limits in the Claude API (Anthropic) service file: at most 16 at a time per connection. Repeated reads are answered from cache (3 Claude API (Anthropic) actions can be), which never reaches Claude API (Anthropic). **What does a Claude API (Anthropic) action cost?** One upstream call, the only unit actionlake bills. A list that pages bills one call per page (113 Claude API (Anthropic) actions page). Cache hits, searches, dry runs and calls Claude API (Anthropic) refused with 401, 403, 429 or a 5xx are free. The Free plan includes 100,000 calls a month with no card. Paid plans start at US$49 a month for 250,000 calls, with overage from US$0.30 down to US$0.18 per 1,000. **Which Claude API (Anthropic) API does it cover?** The Claude API (Anthropic) API 2023-06-01 at api.anthropic.com. It is built from Claude API (Anthropic)'s documentation as read on 1 October 2026. 113 actions in 27 areas, each one in the signed service file (version 1.0.0). **Does actionlake keep my Claude API (Anthropic) data, and where?** actionlake processes and stores data in Sydney, Australia; there is no US region yet. Reads may be cached for up to an hour, and a write clears the reads it affects. Each run's request and response is kept, sealed, for your plan's trace window: 1 day on Free, 3 days on Starter, 7 days on Pro, 14 days on Team, 30 days on Business and 30 days on Enterprise. Your Claude API (Anthropic) credential is sealed under your tenant's key, never logged and never returned. Your traffic is not used to train models. **Which AI clients can use it?** Claude Code, tested by us end to end, and the Claude app as a custom connector, tested by us with an existing key. Other MCP clients that support remote servers over Streamable HTTP (MCP revisions 2025-06-18 and 2025-11-25) and OAuth dynamic client registration, or a bearer header, should connect the same way; we have not tested them. Or plain REST, under the same action names. ## More connectors like Claude API (Anthropic) Shares the most action names with - [OpenRouter 16](https://actionlake.com/connectors/openrouter) - [OpenAI Platform 16](https://actionlake.com/connectors/openai-platform) - [Google Gemini API (AI Studio) 11](https://actionlake.com/connectors/google-gemini-api-ai-studio) - [Groq 9](https://actionlake.com/connectors/groq) - [Perplexity 9](https://actionlake.com/connectors/perplexity) - [xAI Grok API 9](https://actionlake.com/connectors/xai-grok-api) Joins with, on shared entities - [Slack 1](https://actionlake.com/connectors/slack) - [Canvas LMS 1](https://actionlake.com/connectors/canvas-lms) - [Unified.to 1](https://actionlake.com/connectors/unified-to) - [Appwrite Cloud 1](https://actionlake.com/connectors/appwrite-cloud) - [Teamwork 1](https://actionlake.com/connectors/teamwork) - [ERPNext (Frappe) 1](https://actionlake.com/connectors/erpnext-frappe) Browse - [All AI connectors](https://actionlake.com/connectors/category/ai) - [All connectors](https://actionlake.com/connectors) - [How actionlake compares](https://actionlake.com/compare) actionlake is independent. It is not made or endorsed by Claude API (Anthropic); the name is the vendor's own. Counts are from the Claude API (Anthropic) service file, version 1.0.0. ## Give your agent Claude API (Anthropic) today. [Connect Claude API (Anthropic) free](https://actionlake.com/console?start=sandbox) [See pricing](https://actionlake.com/pricing) Start free - 100,000 calls/month, no card. No signup. Hard caps on every key. Cached answers are free. --- Source: https://actionlake.com/connectors/anthropic-claude-api More: https://actionlake.com/llms.txt