This example is the deterministic thesis case study. It demonstrates a trusted Python source that turns project notes into a typed report object without using remote OAuth, LLM calls, or provider quota.

Files

input.md — fixture notes.
cap-input.json — capability-call payload generated from the fixture notes.
run-input.json — workflow-run payload containing the fixture notes by value.
workflow.plan.json — three-node raw plan that passes notes by value: read_notes -> extract_report -> render_markdown_report.
ops.py — Python source exposing read_notes, extract_report, and render_markdown_report.
wf.config.json — local server/client config using the local.report Python source.

Run

From the repository root:

uv run wf config validate examples/report_workflow/wf.config.json
uv run wf-rpc-server --config examples/report_workflow/wf.config.json

In another terminal:

uv run wf --config examples/report_workflow/wf.config.json status
uv run wf --config examples/report_workflow/wf.config.json cap list --source local.report
uv run wf --config examples/report_workflow/wf.config.json cap call local.report.extract_report --input-file examples/report_workflow/cap-input.json --format compact

The full artifact/deployment/run path is covered by tests/examples/test_report_workflow_example.py. To exercise the same three-node lifecycle manually through the CLI, import the raw plan:

uv run wf --config examples/report_workflow/wf.config.json artifact create-from-plan examples/report_workflow/workflow.plan.json --artifact report_case_study --version 1 --title "Report Case Study" --outcome ok
uv run wf --config examples/report_workflow/wf.config.json deploy save report_case_study.default --artifact report_case_study --version 1 --binding local.report=local.report
uv run wf --config examples/report_workflow/wf.config.json deploy validate report_case_study.default
uv run wf --config examples/report_workflow/wf.config.json run start report_case_study.default --input-file examples/report_workflow/run-input.json --trace-from 0 --trace-limit 5
uv run wf --config examples/report_workflow/wf.config.json run list --limit 5
uv run wf --config examples/report_workflow/wf.config.json run inspect <run_id>
uv run wf --config examples/report_workflow/wf.config.json run trace <run_id> --from 0 --limit 5

Draft workspaces are still useful when an agent starts from one capability and edits toward a complete workflow:

uv run wf --config examples/report_workflow/wf.config.json draft create-from-capability report_ws local.report.extract_report --name report_case_study --title "Report Case Study"
uv run wf --config examples/report_workflow/wf.config.json draft set-name report_ws --revision 1 --name report_case_study
uv run wf --config examples/report_workflow/wf.config.json draft set-input report_ws --revision 2 --step call --map input.text=text
uv run wf --config examples/report_workflow/wf.config.json draft set-output report_ws --revision 3 --step call --map title=state.title --map summary=state.summary
uv run wf --config examples/report_workflow/wf.config.json draft validate report_ws

draft create-from-capability is a best-effort bootstrapper. Focused commands cover common edits to existing draft fields; structural edits such as adding the read_notes and render_markdown_report steps use draft patch, or the raw plan import shown above.

The expected report includes:

title: Weekly Project Update
three action items
at least one risk mentioning Google Drive MCP quota
followups for Markdown rendering and baseline comparison
rendered Markdown beginning with # Weekly Project Update

read_notes still accepts a relative path for compatibility, but workflow runs should prefer the by-value text input. This keeps challenge workspaces self-contained and avoids source-directory path resolution surprises.

Thesis Evidence

The example supports these claims:

Python sources can expose typed capabilities through the same workflow surface as built-in and MCP sources.
The case-study path is deterministic and does not depend on an LLM or remote provider.
The workflow lifecycle can be exercised through config validation, capability inventory, capability calls, artifacts, deployments, runs, inspect, and trace.