Drive the actual Strands loop with a local model

Stage 4 of 4. Read the project prerequisites before starting; this stage builds on the preceding contracts.

What changes

The real Strands Agent consumes streaming events from an injected Model subclass. This exercises the framework loop without credentials or cloud calls. The model still only proposes a plan; parse and scope-check its result with the earlier validator. A separate Bedrock constructor is explicit and is not invoked by offline demos or tests.

Work through one concrete case

The optional Strands Agent emits a JSON plan through its real streaming Model interface. That proposed plan then passes through the same scope validator, retry wrapper and bounded executor used by cli.py.

Change the lab inputs and calculate the result before reading its metrics. The figure computes from those inputs; the implementation tests below remain the source of completion evidence.

Implement the contract

Implement strands_adapter.py: parse_model_plan, run_strands, bedrock_agent. This artifact is stage 4 of Cloud Agent With AWS Strands. It consumes explicit inputs and returns an inspectable result that the next stage can use.

Use the public API contract and the typed starter signatures. Return values from core functions and let the supplied driver own file input, argument parsing and presentation.

Keep three modes explicit: pure offline core, installed-SDK local-model comparison, and opt-in AWS reads. An installed SDK passing a fake-model test is useful integration evidence but says nothing about IAM or Bedrock availability.

Verify and inspect

From the repository root, initialize once with python3 scripts/project_test.py cloud-agent-with-aws-strands --init learning-artifacts/cloud-agent-with-aws-strands. Then grade cumulatively:

python3 scripts/project_test.py cloud-agent-with-aws-strands --stage 4 --path learning-artifacts/cloud-agent-with-aws-strands --strict

A fresh workspace should fail until you implement the contract. After every stage is complete, run your actual artifact from the supplied sample:

cd learning-artifacts/cloud-agent-with-aws-strands
python3 cli.py samples/input.json --output incident.json

Investigate the failure boundary

Run the sample through cli.py and inspect cached=true on the repeated metrics read. Then run --optional --strict with the pinned SDK to exercise framework events without sending cloud requests.

Default mode reads the supplied recording. --mode aws is an opt-in AWS CLI adapter for ECS services, CloudWatch CPU and bounded log reads; it requires caller configuration, credentials and AWS permissions. No deployment or live verification is implied.

Verify the actual framework

The five default stage tests check the adapter contract without importing the SDK. Install the pinned optional dependency, then include the five real-SDK tests explicitly:

python3 -m venv .venv
.venv/bin/python -m pip install -r projects/cloud-agent-with-aws-strands/requirements-framework.txt
.venv/bin/python scripts/project_test.py cloud-agent-with-aws-strands --solution --optional --strict
.venv/bin/python projects/cloud-agent-with-aws-strands/solution/framework_demo.py

Use --path my-cloud-agent-with-aws-strands instead of --solution to grade your implementation. Missing dependencies produce a skip in optional mode and a failure in strict optional mode. The verified SDK version is strands-agents==1.57.1; all model replies are local fixtures.

References

Reference 1