Skip to content

DEMOS.md

Latest commit

 

History

History
39 lines (30 loc) · 2.57 KB

DEMOS.md

File metadata and controls

39 lines (30 loc) · 2.57 KB
Raw

Demos

Five runnable scenarios in ../demos/, each targeting a different audience. Every scenario drives the real cyclework API (Engine, Verdict, the Result/Iteration trace), touches no network, prints narrated output, and exits 0 — so they double as smoke tests of the public surface.

python demos/run_all.py               # all five, end to end
python demos/02_numeric_solvers.py    # or just one

On a Windows / cp1252 console, run with PYTHONUTF8=1 so the output encodes cleanly.

# Demo Audience Shows
1 01_agent_retry_loop.py AI / LLM agent builders A self-correcting "fix until tests pass" loop as a first-class, inspectable object — check reports one blocker, revise fixes exactly it, and the full trace is the audit trail.
2 02_numeric_solvers.py Scientific / numerical engineers Convergence as a refinement loop: Newton's sqrt, the cos(x) fixed point, and a geometric series built inline — three problems, one engine.
3 03_plateau_and_budget.py SREs / production loop owners Honest stopping: the engine driven into all four terminal states (SOLVED / PLATEAU / EXHAUSTED / ERROR) on purpose, always keeping the best candidate seen.
4 04_feedback_refiner.py Data / content pipeline engineers Feedback-driven revision: Verdict.feedback names the first broken rule and revise applies exactly that fix, normalizing messy text into a URL slug one cycle at a time.
5 05_streaming_observability.py Observability / platform engineers Watching a loop live: the on_iteration push callback (a live progress bar) and the Engine.cycle() pull generator that lets the caller stop early.

The shape they share

Every demo is the same loop with a different check/revise pair:

  1. check(state) -> Verdict — score the candidate, decide if it's good enough, and (optionally) hand feedback to the next step.
  2. revise(state, verdict) -> state — produce the next candidate, using the feedback when there is some.
  3. The engine records each step, detects convergence/plateau, enforces the budget, and returns a Result with the full trace.

Each demo prints clear, narrated output and exits 0, so they double as smoke tests — tests/test_demos.py runs every one under pytest, alongside the unit tests for the engine and examples.