Registry / llm-agents / ibm-watsonx-orchestrate

ibm-watsonx-orchestrate

JSON →
library2.11.0pypypiunverified

The IBM watsonx Orchestrate Python SDK (ADK) enables developers to build, orchestrate, and deploy AI assistants and workflows using watsonx Orchestrate capabilities. Current version is 2.9.0, with a rapid release cadence. Requires Python >=3.11,<3.14.

pip install ibm-watsonx-orchestrate
INSTALL
IMPORT
SIG · IBM-WATSONX-ORCHES
I
ibm-watsonx-orchestrate
llm-agentspythonv2.11.0
Install
9.1s avg
Import
Disk
148MB
Pass rate
6/ 10
Env Coverage6 / 10
glibc
3.93.13
musl
3.93.13
Install & Compatibility
Where this runs
tested against v2.11.0 · pip install
no network on importno background threads
Install × environment matrix
Each cell = how many times install + import succeeded across repeated harness runs. Partial = flaky.
glibc = Debian/Ubuntu slim · musl = Alpine Linux
musl
glibc
py 3.10
✕ build_error
✕ build_error
py 3.11
✓ —
✓ 10.1s
py 3.12
✓ —
✓ 8.6s
py 3.13
✓ —
✓ 8.7s
py 3.9
✕ build_error
✕ build_error
148MB installed
● package 148MB
Code
Verified usage

Verified import paths — ran on the pinned version, not inferred.

Orchestrate
from ibm_watsonx_orchestrate import Orchestrate
from ibm_watsonx_orchestrate import Orchestrate

Initialize the SDK, create a skill with a simple function, and deploy it to watsonx Orchestrate. Requires WATSONX_API_KEY and WATSONX_URL environment variables.

from ibm_watsonx_orchestrate import Orchestrate, Skill orchestrate = Orchestrate( api_key=os.environ.get('WATSONX_API_KEY', ''), url=os.environ.get('WATSONX_URL', '') ) # Create a skill skill = Skill( name='HelloSkill', description='A simple greeting skill', actions=[ { 'name': 'greet', 'description': 'Greets the user', 'type': 'function', 'function': { 'parameters': { 'type': 'object', 'properties': { 'name': {'type': 'string'} } } } } ] ) # Deploy the skill response = orchestrate.skills.create(skill) print(f'Skill created: {response['result']['id']}') print('Quickstart complete!')
Debug
Known issues
breakingIn v2.0, the SDK was restructured: imports changed from `ibm_watsonx_orchestrate.api` to `ibm_watsonx_orchestrate`. The `Orchestrate` client no longer accepts `iam_apikey`; use `api_key` instead.
fix
Update imports: `from ibm_watsonx_orchestrate import Orchestrate`. Replace `iam_apikey` parameter with `api_key`.
affects: <=1.x -> 2.0+
deprecatedMethod `orchestrate.skills.list()` is deprecated in favor of `orchestrate.skills.search()`.
fix
Replace `list()` with `search()` and adjust parameters if needed.
affects: >=2.5.0
gotchaThe SDK does not automatically retry on rate limits (429 errors). You must implement your own retry logic or use an HTTP client with backoff.
fix
Wrap calls with `tenacity` or `backoff` to handle retries.
affects: all
Upgrade
Version history
2.11.0latest on PyPI · released Jun 13, 2026
Audit
Dependencies
ibm-cloud-sdk-corerequiredCore authentication and HTTP client for IBM Cloud services
ibm-watsonx-airequiredRequired for invoking watsonx.ai models within skills
Agent activity
25 hits · last 30 days
node
24
OpenAI (training)
1
Resources
ibm-watsonx-orchestrate — pip install ibm-watsonx-orchestrate · libregistry