List Agent Tools
List the tools attached to an agent in a simple, unified shape.
Request
- Base URLhttps://api.xpander.ai (the document declares no server; this is the origin it was published from)
- AuthAPI key in header x-api-key
Path parameters
Response
Successful Response
Stable handle for this attached tool. Pass it to DELETE to remove the tool.
Type of a tool attached to an agent.
Human-readable tool name.
Connector connection id. Actions only.
Operation catalog id. Actions only.
OpenAPI operationId. Actions only.
Operation serves AI-generated mock data instead of the real API. Actions only.
Binding auth mode: 'autonomous' (connection credential) or 'user_authentication' (each caller signs in). Actions only.
Whether a user's authorized connector token is shared across the org's other agents. Actions only.
What happens to a call that changes something: 'allow', 'ask' (a person approves it first) or 'block'. Actions only; None means allow.
Whether a person must approve this action before it runs. Actions only; None means never configured.
The stored approval rule for this action (approvers, channels, allowed scopes, on_deny), so a settings UI opens on real values. Never carries arguments or credentials.
Custom function id. Custom functions only.
MCP registry id. MCP only.
Subset of MCP tools exposed. MCP only.
MCP server url. MCP only.
MCP server type: remote (url) or local (command, hosted in the agent's workspace). MCP only.
Command that starts a local MCP server. Local MCP only.
Environment of a local MCP server. A ${NAME} reference to a workspace secret is returned as written; a literal value is returned blank, the same way header values are. Local MCP only.
MCP auth type (api_key/oauth2/custom_headers/sso_user/none). MCP only.
Registry entry id when this MCP is attached from the shared org registry (editing it affects every agent). Absent for inline MCPs. MCP only.
MCP transport (streamable-http/sse/stdio). MCP only.
Name of the bound registry entry, read for every viewer, also when the entry is another member's personal one. Set by the agent tools list (GET /v1/agents/{id}/tools), not by the agent views. MCP only.
Audience of the bound registry entry: personal or organizational. MCP only.
Display name of the member who owns a personal registry entry. MCP only.
Custom header names configured on the MCP (keys only; values are never returned). MCP only.
OAuth2 client id for the MCP, if set. MCP only.
Extra OAuth2 scopes for the MCP. MCP only.
Whether the authorized MCP token is shared across the org's other agents. MCP only.
Whether a caller with no sign-in of their own runs this agent's OAuth MCP on the owner's connection. MCP only.
For mcp_auth_type sso_user: access, id or exchange. MCP only.
For mcp_auth_type sso_user with exchange: the audience. MCP only.
Whether an API key is stored for the MCP (the value is never returned). MCP only.
Whether an OAuth2 client secret is stored for the MCP (the value is never returned). MCP only.
Last probe outcome for the MCP: connected / tools_failing / auth_required / error. MCP only.
True when the last probe's health-check tool call succeeded, so the tools are known to run. MCP only.
Why the last probe did not report connected. MCP only.
The tool the health probe calls to prove calls work. MCP only.
Text that marks a failed health-check answer. MCP only.
Tool count discovered by the last probe. MCP only.
ISO timestamp of the last probe. MCP only.
Referenced sub-agent or workflow id.
Skill name from the xpander skills catalog. Skills only.
False when the tool is authored hidden and must not be listed in the app; it stays attached and keeps running.
Changes
No recorded changes to this operation across all 5 revisions of this API.