← Instrument galleryD05-ACTOR-RUNTIME-THEATER

Actor Runtime Theater

A deterministic browser-local research instrument.

DETERMINISTIC FIXTURE · NOT A PERFORMANCE RESULT

Actor Runtime Theater

MECHANISMPersistent Actor capsules with temporary Agent processes
AUTHORITYActor revision and validated world state
COMMIT RULEOnly accepted World Patch operations increment revision.
FAIL-CLOSED RULEUnknown facts and direct capsule writes are excluded.
STATEDORMANT597361a8
actor revision0
knowledge scopeDORMANT
model calls597361a8
  1. DORMANT
  2. ELIGIBLE
  3. SELECTED
  4. ACTIVE
  5. PROPOSED
  6. COMMITTED
  7. COOLDOWN
CONTRACT CONTROLS

Each control executes a declared browser-local transition. Repeated dual-outcome controls alternate deterministically.

FIXTURE ASSERTIONS

NOT YET EXERCISEDDormant actors consume no model call.

NOT YET EXERCISEDAgent packet excludes unknown world facts.

NOT YET EXERCISEDAgent cannot directly mutate Actor Capsule.

NOT YET EXERCISEDAccepted operations increment revision and preserve provenance.

RECEIPTED EVENT LOG

No transition yet. Operate a declared control.

This transparent browser-local instrument teaches the authored state contract. Exercising an assertion demonstrates fixture behavior only; it is not evidence of production performance, compression ratio, model uplift, cognition, safety, or universal correctness.

Purpose

Show a durable Actor Capsule waking into a temporary Agent Process and returning a bounded World Patch.

Mechanism

A deterministic scheduler scores actors. A selected actor receives limited knowledge, goals, permissions, and context. The temporary process proposes beliefs, mood, relationships, schedule, and actions. Validators apply bounds and provenance before a revision commits.

Deterministic fixture

Five city actors with different goals, relationships, beliefs, deadlines, cooldowns, and one false belief that is valid in actor state.

The fixture must live in content/demos/d05-actor-runtime-theater/fixture.json, validate before the demo renders, and be resettable without a network call. A future live adapter may be added behind a visibly separate mode.

Required controls

  • Advance time block
  • Trigger event
  • Inspect score
  • Select actor
  • Run multiplex pass
  • Edit context budget
  • Inspect World Patch
  • Accept/reject operations
  • Replay revision

Visible states

  • DORMANT
  • ELIGIBLE
  • SELECTED
  • ACTIVE
  • PROPOSED
  • COMMITTED
  • COOLDOWN

Every state must have text, iconography, and a non-color-only distinction. State transitions are logged in an in-memory demo receipt visible in the inspector.

Core assertions

  1. Dormant actors consume no model call.
  2. Agent packet excludes unknown world facts.
  3. Agent cannot directly mutate Actor Capsule.
  4. Accepted operations increment revision and preserve provenance.

Layout

The page contains:

  1. a concise mechanism explanation;
  2. the interactive stage;
  3. a state and evidence inspector;
  4. a reset control;
  5. a “what this proves / what it does not prove” panel;
  6. links to the related paper and benchmark contract.

On narrow screens the inspector becomes a bottom sheet. The stage must remain usable at 360 CSS pixels.

Accessibility

  • All operations are keyboard reachable.
  • Pointer gestures have button or keyboard equivalents.
  • Animated transitions honor prefers-reduced-motion.
  • Focus remains visible and returns predictably after sheets close.
  • Diagrams expose concise text alternatives.
  • Status is announced through a polite live region only when the user initiates the transition.

Instrumentation

Record locally:

demo_id
fixture_version
action
prior_state
result_state
validation_result
timestamp_relative

Do not send telemetry. Provide “Export demo trace” as JSON for debugging and reproducibility.

Automated tests

  • fixture schema validation;
  • deterministic reset;
  • every required control transition;
  • every assertion above;
  • invalid and stale fixture handling;
  • keyboard path;
  • 360 px and 1440 px screenshots;
  • zero critical console errors;
  • no request to a paid API in default mode.

Publication boundary

Dialogue and planning may use deterministic fixtures. The demo does not claim a live large-population LLM simulation.

The interface must never convert illustrative values into a benchmark claim. A measured result appears only when linked to a versioned run receipt.