Skip to main content

marbots-sdk (Python)

Typed, zero-dependency Python client for Marbots, the multi-agent collaboration platform. Responses are frozen dataclasses; kernel packs, permission profiles, event types and model settings are enums/helpers, so mypy (and your IDE) catch typos before you run anything.

pip install marbots-sdk
from marbots_sdk import MarbotsClient, BotSpec, KernelPack, PermissionProfile, ModelRef, EventType

mb = MarbotsClient("http://localhost:5170")            # api_key="..." if the server requires one

# Every bot can use its own model; ModelRef.DEFAULT follows the workspace default.
sari = mb.bots.create(BotSpec(
    name="Sari", role="UX designer",
    kernel_functions=[KernelPack.FILES, KernelPack.WEB],
    permission_profile=PermissionProfile.WORKSPACE_WRITE,
    model=ModelRef.of("azure", "gpt-5.6-luna"),
))
mb.bots.set_model("atlas", ModelRef.DEFAULT)
print(mb.models.list().default)                         # e.g. azure/gpt-5-mini

thread = mb.threads.create(sari.id)
result = mb.threads.send(thread.id, "Sketch a landing-page wireframe", wait=True)
print(result.task.model, result.task.cost_usd, result.text)

for event in mb.events.stream(thread.id):               # live Server-Sent Events
    if event.type == EventType.TOOL_CALL_STARTED:
        print("tool:", event.message)
    if event.is_task_finished():
        break
Area API
Bots bots.list/get/create(BotSpec)/update/delete/hire(template)/get_model/set_model/pause/resume/export/import_package
Models models.list() → ModelCatalog, models.set_default("provider/model")
Chat threads.create/send(wait=True)/messages/files/download/export_transcript, chat(bot, text)
Work tasks, approvals.approve(id, scope="Session"), schedules.create(ScheduleSpec), memory, skills, mcp
Live events.stream(thread_id) → AgentEvent

Type safety is tested: python tests/typecheck/run.py requires mypy --strict to pass on the SDK and to reject every deliberate typo in tests/typecheck/check_types.py. The conformance tests (tests/test_conformance.py) run against a real Marbots server.

Built by Gravicode Studios, led by Kang Fadhil. MIT license.

Metadata

Release files for marbots-sdk 0.1.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for marbots-sdk 0.1.0
File Size Uploaded Provenance Actions
marbots_sdk-0.1.0.tar.gz 15.0 kB Without provenance Details for marbots_sdk-0.1.0.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for marbots-sdk 0.1.0
File Interpreter ABI Platform Actions
marbots_sdk-0.1.0-py3-none-any.whl Python 3 none any Details for marbots_sdk-0.1.0-py3-none-any.whl

Total release size: 28.1 kB

Release files / marbots_sdk-0.1.0.tar.gz

Download URL marbots_sdk-0.1.0.tar.gz
Size 15.0 kB
Tags Source
SHA-256 checksum
How to use checksums
bfc76471a67338a0cc5b657c68b53718efc856b9b34b0d64a4738b67917fc5ec
BLAKE2b-256 checksum
How to use checksums
db07ee1724cc10c7ac3fafb6326bd98806530a62abf5fefffadfeee80e607215
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.12.10

Release files / marbots_sdk-0.1.0-py3-none-any.whl

Download URL marbots_sdk-0.1.0-py3-none-any.whl
Size 13.2 kB
Tags Python 3
SHA-256 checksum
How to use checksums
77ac318fc5f44d9707e4ecfaf684aa54f9ba174b41b09a523030c84294b78fcd
BLAKE2b-256 checksum
How to use checksums
7ea1bdf202c9982802add8db52002dc6dc69cd7feb1bbaa6e8dbed9a07e9c485
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.12.10

Security for marbots-sdk 0.1.0

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page