Account & control plane

Python SDK reference for identity, usage, API keys, teams, billing, devices, audit logs, execution policy, and the platform contract — every AlgentaClient method and endpoint.

These AlgentaClient methods manage who you are and how your organization is governed: identity and usage, API keys, team members, billing and metering, devices, audit logs, execution policy, and the machine-readable platform contract. Each is a faithful client over the /v1 HTTP contract; the async client AsyncAlgentaClient exposes the same names to await.

import os
from decision_engine import AlgentaClient

client = AlgentaClient(
    api_key=os.environ["ALGENTA_API_KEY"],
    base_url="https://api.algenta.ai",   # or http://localhost:8000 self-hosted
)

Identity and usage

Method	HTTP	Returns
me()	GET /v1/me	MeResult
update_me(*, name=None, org_name=None)	PATCH /v1/me	Any
usage()	GET /v1/usage	UsageInfo
limits()	GET /v1/limits	dict
distributions()	GET /v1/distributions	DistributionListResult
templates()	GET /v1/templates	TemplateListResult
health()	GET /v1/health	dict
version()	GET /v1/version	dict

• me returns your user and organization (me.user, me.org); update_me renames your user (name) or organization (org_name).
• usage reports the current billing_period, simulations_run, api_calls, and quota_used_pct; limits returns the raw quota ceilings.
• health and version are unauthenticated liveness and build endpoints.

API keys

Method	HTTP	Returns
list_api_keys()	GET /v1/api-keys	list[APIKeyInfo]
create_api_key(label=None, *, name=None, expires_at=None, device_limit=None)	POST /v1/api-keys	dict
revoke_api_key(key_id)	DELETE /v1/api-keys/{key_id}	dict

• create_api_key returns the new key's full secret exactly once — store it immediately. device_limit caps how many devices may bind to the key; expires_at accepts a datetime or ISO-8601 string.
• list_api_keys returns metadata only (id, label, key_prefix, status), never the secret.

Team members

Method	HTTP	Returns
list_team_members(*, page=None, limit=None)	GET /v1/team	TeamListResult
invite_team_member(*, email, role="member")	POST /v1/team/invite	TeamInviteResult
update_team_member_role(user_id, *, role)	PATCH /v1/team/{user_id}/role	TeamRoleUpdateResult
remove_team_member(user_id)	DELETE /v1/team/{user_id}	TeamRemoveResult

Billing, credits and metering

Method	HTTP	Returns
get_billing_info()	GET /v1/billing/info	BillingInfoResult
create_billing_checkout(*, plan=None)	POST /v1/billing/checkout	BillingSessionResult
create_billing_portal()	POST /v1/billing/portal	BillingSessionResult
refresh_credits(*, device_id, billing_period, credits_used=0)	POST /v1/credits/refresh	CreditRefreshResult
ingest_metering_events(*, device_id, events)	POST /v1/metering	MeteringBatchResult

• create_billing_checkout and create_billing_portal return a hosted session you redirect the user to.
• refresh_credits and ingest_metering_events are generic managed-runtime compatibility methods. Self-hosted Algenta does not call them for billing or enforcement; decisions are unlimited and audit records remain local.

Devices

Method	HTTP	Returns
list_devices(*, page=None, limit=None)	GET /v1/device/list	DeviceListResult
revoke_device(registration_id)	DELETE /v1/device/{registration_id}	DeviceRevokeResult

Audit logs and execution policy

Method	HTTP	Returns
get_audit_logs(*, page=1, limit=25, actor_email=None, action=None, resource_type=None, result=None, policy_snapshot_id=None, schema_snapshot_id=None, manifest_version=None, request_hash=None)	GET /v1/audit-logs	AuditLogResult
get_audit_log_artifacts(*, page=1, limit=25, actor_email=None, action=None, resource_type=None, result=None, policy_snapshot_id=None, schema_snapshot_id=None, manifest_version=None, request_hash=None, content_hash=None)	GET /v1/audit-logs/artifacts	Any
get_execution_policy()	GET /v1/execution/policy	ExecutionPolicyResult
list_execution_policy_snapshots()	GET /v1/execution/policy/snapshots	ExecutionPolicySnapshotListResult
update_execution_policy(*, min_confidence=None, risk_floor=None, require_calibration=None, allow_reexecution=None)	PATCH /v1/execution/policy	ExecutionPolicyResult

• Audit queries filter on provenance (policy_snapshot_id, schema_snapshot_id, manifest_version, request_hash) so you can reconstruct exactly what governed any action; get_audit_log_artifacts additionally filters by content_hash.
• update_execution_policy tunes the gate that decides whether a recommendation may auto-execute: min_confidence, risk_floor, require_calibration, and allow_reexecution. Only the fields you pass are changed.

Platform contract and runtime

Method	HTTP	Returns
get_contract()	GET /v1/meta/contract	PlatformContractResult
get_runtime_manifest()	GET /v1/runtime/manifest	RuntimeManifestResult
get_runtime_modules()	GET /v1/admin/runtime/modules	RuntimeAdminModulesResult
get_runtime_benchmarks()	GET /v1/admin/runtime/benchmarks	RuntimeAdminBenchmarksResult
get_runtime_release_validation()	GET /v1/admin/runtime/validation	RuntimeReleaseValidationResult

• get_contract is the machine-readable description of everything this instance supports — endpoint families, SDK/CLI/MCP method names, and governed-filter rules. It is what the SDKs bootstrap from.
• get_runtime_manifest reports the signed engine build serving requests; the admin/runtime/* methods surface module inventory, benchmarks, and release validation for the running engine.

Every list result carries pagination fields (total, page, limit, pages); page until you have what you need. Snapshot ids (policy_snapshot_id, schema_snapshot_id) returned by these methods are the same values you can pass as filters to get_audit_logs and the agent-run query methods.

Example: who am I, and how much have I used

import os
from decision_engine import AlgentaClient

client = AlgentaClient(api_key=os.environ["ALGENTA_API_KEY"], base_url="https://api.algenta.ai")

me = client.me()
print(me.user.email, "in", me.org.name, f"({me.org.plan})")

usage = client.usage()
print(f"{usage.billing_period}: {usage.api_calls} API calls, {usage.quota_used_pct:.1f}% of quota")

for key in client.list_api_keys():
    print(key.key_prefix, key.label, key.status)

Expected output resembles:

[e-mail] in Acme (team)
2026-07: 1842 API calls, 18.4% of quota
de_live_ab12 CI key active

Example: tighten the execution policy

import os
from decision_engine import AlgentaClient

client = AlgentaClient(api_key=os.environ["ALGENTA_API_KEY"], base_url="https://api.algenta.ai")

current = client.get_execution_policy()
print("min_confidence before:", current.min_confidence)

updated = client.update_execution_policy(
    min_confidence=0.9,
    require_calibration=True,
)
print("min_confidence after:", updated.min_confidence)

# Every policy change is recorded — read it back from the audit log.
logs = client.get_audit_logs(action="execution_policy.update", limit=5)

Related references

Agent runsAPI overviewUsage & limits