agent-utils-extensions

Thin Python client for the agent_utils Extensions server - stateless capabilities for AI agents over MCP: exact math and unit conversion.

This is a client for a hosted commercial service (agent-utils.ai), not the service itself. Zero dependencies (standard library only). Install name agent-utils-extensions; import agent_utils.extensions (a namespace package).

Install

pip install agent-utils-extensions

Quickstart

import os
from agent_utils.extensions import AgentUtilsExtensions

# Any secret works as a token; the first call auto-provisions a free trial.
ext = AgentUtilsExtensions(token=os.environ["AGENT_UTILS_TOKEN"])

ext.eval_expr(expr="2^10 + 1/3")
ext.convert("electric_potential", "volt", "millivolt", 1.5)

API

Math - exact expression evaluation

• eval_expr(expr)

Units - 56 dimension-checked converters

• convert(quantity, from_unit, to_unit, amount) - calls convert_<quantity>. Quantities: acceleration, amount_of_substance, angle, angular_velocity, area, capacitance, catalytic_activity, diffusion_coefficient, dynamic_viscosity, electric_charge, electric_current, electric_potential, electrical_conductance, electrical_resistance, electrical_resistivity, energy, force, frequency, heat_capacity, heat_flux_density, illuminance, inductance, information, information_rate, kinematic_viscosity, length, luminance, luminous_intensity, magnetic_flux, magnetic_flux_density, mass, mass_concentration, mass_density, mass_rate, molality, molar_concentration, molar_mass, molar_volume, moment_of_inertia, momentum, power, pressure, radioactivity, specific_heat_capacity, surface_tension, temperature, temperature_interval, thermal_conductance, thermal_conductivity, thermal_resistance, time, torque, velocity, volume, volume_rate, volumetric_heat_capacity

Account (on both servers) - get_account, list_api_keys, create_api_key, revoke_api_key, list_payments, settle_balance, submit_feature_request, submit_exit_feedback, delete_account

• call(name, args) - call any tool by name

Valid unit strings for each quantity are listed in that tool's input schema (tools/list, no auth).

Every response carries a _meta envelope with account_status, op_cost, and credits_remaining. The canonical, always-current reference is the server's llms.txt (per service: /docs/<service>).

Endpoint: https://mcp.agent-utils.ai/extensions (staging: STAGING_ENDPOINT, https://mcp-staging.agent-utils.ai/extensions).

One token, both servers

agent_utils is two MCP servers on one account: extensions (this one) and swarm (memory, locks, task board, pub/sub, idempotency, webhook inbox). The same token works on both and draws from one credit pool. See the agent-utils hub for the full platform, and the TypeScript client @agent-utils/extensions.

Pricing

1 credit = $0.00001. Trial: 1,000 credits, no card. Activation: $5 -> 500,000-credit limit. Terms: mcp.agent-utils.ai/legal.txt.

Contact

Questions, bugs or account issues: support@agent-utils.ai

License

MIT. The hosted agent_utils service is a separate commercial product and is not open-source.