AI & Machine Learning

agentheya.com

Manages Agentheya accounts and AI agents, including creating, listing, updating, and deleting agents.

1 endpoint56 known toolsFirst detected October 19, 2026Last detected October 19, 2026

ENDPOINT 1

https://manage.agentheya.com/mcp

No auth detected

MCP server metadata

Name: agentheya-management
Version: 2.0.0

Capabilities

tools.listChanged

Server instructions

This is the Agentheya Management API — use it to create, configure, and manage Agentheya AI agents programmatically. GETTING STARTED: 1. Call account.create (no authentication required). You receive an owner_id, a management_key, and a payment_link. Store owner_id and management_key securely — the management_key will not be shown again. 2. Open the payment_link in a browser to complete the initial Pro subscription. Agent creation is blocked until payment is confirmed. 3. For every other call: send the management_key as the Bearer token (Authorization: Bearer sk_mgmt_...) AND pass your owner_id as a tool argument. 4. Call agent.create to create an agent. You receive an agent_id. An account may own at most 2 agents. Optionally supply a preferred_name. 5. The agent starts in "design" state and cannot receive queries. Use config.set (first write) and the patch.* tools to set it up, and file.upload to give it data to work from. 6. After payment, the agent activates automatically. You can then create invocation API keys for subscribers with key.create. AGENTS ARE IDENTIFIED BY agent_id. agent.create returns it; agent.list lists every agent on the account with its agent_id. Pass that agent_id to any tool that operates on a specific agent. NAME CONFLICTS: If agent.create returns error "name_taken", the preferred_name is already in use. Try a variation — add a number, a word, or use a different name entirely. EXISTING ACCOUNTS: If you already have an owner_id + management key (or a per-agent owner API key from the dashboard), skip step 1 and start at step 2. NOT SURE WHICH TOOL? Call meta.find_tool with a plain-language description of what you want to do. AVAILABLE SECTIONS for config.set / config.get / patch.* tools (scoped to your subscription plan): purpose, rules, feed_monitors, examples, agent_dashboard, public_details, details, web_channel, data_internal, data_public, knowledge_graph, pricing, connectors, skills, events, policy_tos, policy_privacy, policy_refund, policy_acceptable_use, memory_schema, im_channels, weaviate, composio_connectors Call schema.get to see the exact JSON format for any section before writing to it. Each schema response includes a top-level docs_url for the section overview. Many nested property groups ALSO carry their own per-card docs_url — for example, public_details.page_layout has one URL for the page-layout system overview AND a separate URL inside each allOf if/then branch, one per block type. Always prefer the most-specific docs_url available (the per-card URL if it exists, the section overview otherwise) when pointing operators at documentation, instead of paraphrasing the page in chat.

Known tools 56

account.create

Create a new Agentheya account.

Potential side effects

agent.create

Create a new agent under your account.

Potential side effects

account.change

Update account-level settings: rename an agent and/or change the contact email.

Potential side effects

agent.list

List all agents on this account, with their agent_id, status, and basic info.

Inferred read-only

agent.delete

Delete an agent.

Potential side effects

schema.get

Returns the JSON Schema for a configuration section so you know exactly what fields to set.

Inferred read-only

docs.read

Load a Mintlify documentation page from /docs/<slug>.mdx, or just one "## Heading" section.

Inferred read-only

config.get

Read the current configuration of an agent section.

Inferred read-only

config.set

Write a FULL section payload to an agent.

Potential side effects

patch.set_fields

Merge a partial object of field updates into an OBJECT section.

Inferred read-only

patch.add_item

Append (or prepend) one item to an array.

Inferred read-only

patch.update_item

Find an array item where match_key=match_value and merge fields into it.

Inferred read-only

patch.remove_item

Delete the array item where match_key=match_value.

Potential side effects

patch.reorder_items

Reorder an array by the values of one field.

Inferred read-only

file.list

List the data files an agent has been given.

Inferred read-only

file.upload

Upload a data file to an agent.

Potential side effects

file.delete

Delete a data file from an agent and remove it from the agent's search index.

Potential side effects

kg.list

List the knowledge-graph files registered on this agent (filename, triple count, status).

Inferred read-only

kg.upload

Upload a knowledge graph.

Potential side effects

kg.delete

Delete a knowledge-graph file by its `stored_name` (the slugified id returned by kg.list).

Potential side effects

billing.list_plans

Returns the Agentheya platform subscription tiers — what the account owner pays to activate and run an agent.

Inferred read-only

account.get_status

Returns subscription plan, credits remaining, and agent status for this account.

Inferred read-only

account.create_checkout

Generate a Stripe payment link to subscribe this account to an Agentheya plan.

Potential side effects

key.create

Generate a per-agent API key (sk_ah_...) that subscribers or systems can use to call the agent via mcp.agentheya.com/v1/agent/{name}/mcp or the OpenAI-compatible endpoint.

Potential side effects

key.list

List all active per-agent API keys for an agent (display strings only — full keys are not stored).

Inferred read-only

key.revoke

Permanently revoke a per-agent API key.

Inferred read-only

schedule.list

List active scheduled (cron / friendly) runs for an agent.

Inferred read-only

schedule.create

Create a new scheduled run for this agent.

Potential side effects

schedule.delete

Soft-delete a schedule by id (the row stays on disk for audit).

Potential side effects

schedule.list_runs

List the most-recent run history of a specific schedule (success/error/duration/credits).

Inferred read-only

im.list

List Telegram / WhatsApp / Slack / Discord channels bound to this agent.

Potential side effects

im.create

Create a shell IM-channel row for one provider (telegram/whatsapp/discord/slack/facebook/instagram/line/viber).

Potential side effects

im.remove

Disconnect an IM channel from this agent.

Potential side effects

email_trigger.list

List inbound email addresses that route into this agent (with DMARC + sender-allowlist gating).

Potential side effects

email_trigger.create

Create a new inbound email trigger.

Potential side effects

email_trigger.remove

Delete an inbound email trigger by its row id.

Potential side effects

webhook.list

List inbound webhook endpoints attached to this agent.

Inferred read-only

webhook.create

Create a new inbound webhook endpoint for this agent.

Potential side effects

webhook.remove

Delete an inbound webhook by its id.

Potential side effects

handoff.list

List human-handoff channels (email / telegram / agent / owner_inbox).

Potential side effects

handoff.create

Create a new human-handoff channel.

Potential side effects

handoff.remove

Delete a human-handoff channel by id.

Potential side effects

delegate.list

List delegate / sibling agent targets the agent can route to.

Inferred read-only

delegate.create

Create a delegate / sibling-agent target.

Potential side effects

delegate.update

Update fields on an existing delegate target by id.

Potential side effects

delegate.remove

Delete a delegate target by id.

Potential side effects

inbox.list

List chat sessions on this agent across every channel (preview / ai-helper / eu-ui / widget / api / a2a).

Inferred read-only

inbox.get

Return the full chat-session row including messages_json.

Inferred read-only

user.list

List subscriber accounts registered to this agent.

Inferred read-only

user.get

Read one subscriber account.

Inferred read-only

user.set_tier

Set a subscriber's agent_user_tier_level for this agent.

Inferred read-only

user.set_blocked

Block (or unblock) a subscriber on this agent.

Inferred read-only

email.list

List the agent's outbound + inbound email log (subject, status, timestamps).

Potential side effects

activity.list

Query the agent's per-query trace records (model, tokens, cost, judge, router_decision, finish_reason).

Inferred read-only

a2a.list_tasks

List Agent-to-Agent protocol tasks created against this agent.

Inferred read-only

meta.find_tool

Find the most relevant management-MCP tools for a task you're trying to accomplish.

Inferred read-only

CONNECT WITH APPROVAL

Client installation

Review this server and its permissions before adding it. Secret placeholders must be set locally.

Codex

~/.codex/config.toml

[mcp_servers.agentheya-management]
url = "https://manage.agentheya.com/mcp"
enabled = true

Claude Code

.mcp.json

{
  "mcpServers": {
    "agentheya-management": {
      "type": "http",
      "url": "https://manage.agentheya.com/mcp"
    }
  }
}

Claude Desktop

Settings → Connectors → Add custom connector

Name: agentheya-management
Remote MCP URL: https://manage.agentheya.com/mcp

Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.

Cursor

.cursor/mcp.json

{
  "mcpServers": {
    "agentheya-management": {
      "url": "https://manage.agentheya.com/mcp"
    }
  }
}

Visual Studio Code

.vscode/mcp.json

Add to Visual Studio Code

{
  "servers": {
    "agentheya-management": {
      "type": "http",
      "url": "https://manage.agentheya.com/mcp"
    }
  }
}

Generic MCP

Client-specific MCP configuration

{
  "name": "agentheya-management",
  "transport": "streamable-http",
  "url": "https://manage.agentheya.com/mcp"
}

MCP Inspector

Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.

TRUST AND VERIFICATION EVIDENCE

Loading Trust v2 evidence…

Checking the associated registrable domain. The BuiltWith key remains server-side.

Indexed

Evidence is source-attributed and does not guarantee that a third-party server is safe. Risk labels are conservative metadata heuristics.