Anthropic MCP integration for AI agents.
Connect AI agents to Anthropic through 4 structured actions, including count message tokens, create message, and get model. Review authentication, inputs, wo…
- ACTION CONTRACTS
- 4
- AUTHENTICATION
- API key
- SOURCE STATUS
- Provider-backed
CATALOG SOURCE REVIEWED AUGUST 23, 2026 / ACTION NAMES AND SCHEMAS DERIVED FROM WORKING MACHINES PROVIDER SOURCE
CAPABILITY PROFILE
What the Anthropic integration exposes
The Anthropic Agent App exposes 4 provider-backed actions for ai work. Its current contract lets an authorized agent count input tokens for an anthropic message request, create a non-streaming anthropic message, and fetch metadata for one anthropic model. These operations are called through Working Machines as typed capabilities rather than through browser navigation or copied UI steps.
A connection uses API key. Before execution, the agent can inspect the selected action, its required fields, declared scopes, and expected output contract. Provider credentials remain inside the Working Machines runtime; the calling agent receives the capability and its structured result, not the underlying secret.
For reliable operation, start with the narrowest action that satisfies the task, resolve stable provider identifiers before changing state, and validate the returned object or status after execution. Availability still depends on the connected Anthropic account, granted provider permissions, workspace policy, region, plan, and upstream API behavior.
VERIFIED ACTION SAMPLE
Real Anthropic capabilities.
Showing 4 of 4 actions. Risk labels are conservative signals based on operation names, not substitutes for provider documentation or runtime policy.
count_message_tokensREADCount input tokens for an Anthropic message request.
INPUTS: model / messages / system / thinking / tool_choice
create_messageWRITECreate a non-streaming Anthropic message.
INPUTS: model / max_tokens / messages / system / metadata
get_modelREADFetch metadata for one Anthropic model.
INPUTS: model_id
list_modelsREADList Anthropic models available to the current API key.
INPUTS: before_id / after_id / limit
INPUT CONTRACTS
Know what the action needs before it runs.
count_message_tokens
modelREQUIRED- The Anthropic model identifier to use.
messagesREQUIRED- The ordered conversation history to count.
systemOPTIONAL- System prompt content.
thinkingOPTIONAL- A JSON object passed through to the Anthropic API.
tool_choiceOPTIONAL- Tool selection strategy for the request.
create_message
modelREQUIRED- The Anthropic model identifier to use.
max_tokensREQUIRED- The maximum number of tokens to generate.
messagesREQUIRED- The ordered conversation history sent to the model.
systemOPTIONAL- System prompt content.
metadataOPTIONAL- Metadata about the request.
get_model
model_idOPTIONAL- The exact Anthropic model identifier to retrieve.
list_models
before_idOPTIONAL- Returns models before this model identifier.
after_idOPTIONAL- Returns models after this model identifier.
limitOPTIONAL- The maximum number of models to return.
PROVIDER-SPECIFIC WORKFLOWS
Jobs this Agent App can support
Inspect Count Message Tokens
Count input tokens for an Anthropic message request. Use this as a bounded discovery step, retain the returned identifier, and avoid expanding the read beyond the task's stated scope.
count_message_tokensControl Create Message
Create a non-streaming Anthropic message. Resolve the target first, present material changes for confirmation, and make retries idempotent where the provider supports it.
create_messageVerify with Get Model
Fetch metadata for one Anthropic model. Compare the returned provider state with the intended outcome and preserve stable IDs or canonical links in the run record.
get_modelSAFETY BOUNDARY
Operate Anthropic with explicit limits
- Authorize Anthropic with API key and grant only the provider access required by the selected actions.
- 1 action is change-capable by name. Confirm the target identity and material parameters before allowing a write.
- Do not infer permission from catalog visibility. Workspace policy, connection identity, and upstream authorization still govern execution.
- No provider scope string is declared on the sampled actions. Verify the connected account's actual permissions in Anthropic rather than assuming unrestricted access.
- After a call, inspect the structured result and execution record before reporting that the Anthropic task completed successfully.
CONNECTION MODEL
API key
Anthropic API key used with the x-api-key header. Create or manage API keys in the Anthropic Console: https://console.anthropic.com/settings/keys.
Working Machines stores provider credentials behind the execution boundary. An agent can use an authorized connection identity, but catalog discovery alone does not reveal OAuth tokens, API keys, or provider secrets.
- READ SIGNALS
- 3
- WRITE SIGNALS
- 1
- HIGH IMPACT
- 0
- REVIEW SIGNALS
- 0
EVIDENCE AND AVAILABILITY
Provider reference
Action names, input fields, authentication types, and counts on this page are generated from the Working Machines provider catalog. Provider behavior, quotas, object semantics, account eligibility, and regional availability remain governed by Anthropic.