core/docs/workbench
Shay 02f02c804a docs(workbench): Wave M plan (mastery & worthiness) + Phase B calibration brief pack
Wave M takes the workbench to mastery and closes its biggest design gap:
it surfaces the teaching/ratification loop but is blind to the
calibrated-learning / serving-discipline loop (gold-tether arena, reliability
gate, Wilson floor vs θ ceiling, 'the engine earns the right to guess') and
to cognition itself (pipeline stages, field substrate, identity continuity).
Lens: Anthropic + xAI as target users who'd WANT to use it.

- wave-M-worthiness.md: full plan, Phases A–E, the missing-surfaces table,
  the backend-reader-first / never-re-implement-engine-math disciplines,
  execution order (B→C→D, A parallel).
- wave-M-phaseB-calibration-briefs: executable Phase B pack grounded in the
  real core/reliability_gate shapes (ClassTally / conservative_floor /
  license_for / Action θ) and the committed report.json evidence — B1
  readers (GATING, Python), B2 Calibration route, B3 wrong=0 global frame,
  B4 leeway wiring. Dependency DAG + STOP gates + no-theater rules.
2026-06-12 23:51:57 -07:00
..
acceptance-gates.md docs(workbench): reconcile mutation doctrine with admitted corridors 2026-06-12 06:50:26 -07:00
api-contract-v1.md feat(workbench): sealed single-turn replay backend — GET /replay/{turn_id} (Wave R3) 2026-06-12 17:15:39 -07:00
core-workbench-v1-blueprint.md
data-shapes-v1.md feat(workbench): add R2 backend read substrate 2026-06-12 12:49:41 -07:00
design-system.md
implementation-plan.md docs(workbench): reconcile mutation doctrine with admitted corridors 2026-06-12 06:50:26 -07:00
README.md
ui-component-map.md
wave-1-evidence-spine.md docs(workbench): Wave R mastery revamp plan + R0/R1 dispatch brief pack 2026-06-12 10:02:46 -07:00
wave-M-worthiness.md docs(workbench): Wave M plan (mastery & worthiness) + Phase B calibration brief pack 2026-06-12 23:51:57 -07:00
wave-R-mastery-revamp.md docs(workbench): R0a CI path filter includes the workflow file itself 2026-06-12 10:29:29 -07:00

CORE Workbench v1

CORE Workbench is the first dedicated operator/auditor UI for CORE.

It is intentionally designed as:

  • a cognition observatory,
  • replay debugger,
  • proposal review workstation,
  • deterministic eval console,
  • and audit surface.

It is NOT intended to be a generic chatbot shell.


Document Index

Doctrine

  • ../decisions/ADR-0160-core-workbench-v1.md

Defines:

  • product doctrine
  • trust boundaries
  • architecture direction
  • module scope
  • acceptance criteria

UX Blueprint

  • core-workbench-v1-blueprint.md

Defines:

  • visual philosophy
  • interaction model
  • module behavior
  • navigation
  • replay/proposal/eval UX

Implementation Plan

  • implementation-plan.md

Defines:

  • W-026 through W-031
  • phase sequencing
  • backend/frontend constraints
  • release criteria

API Contract

  • api-contract-v1.md

Defines:

  • read-only endpoint surface
  • request/response envelopes
  • mutation boundaries
  • runtime/trace/proposal/eval/replay routes

Data Shapes

  • data-shapes-v1.md

Defines:

  • typed UI-facing shapes
  • proposal structures
  • replay structures
  • artifact references
  • trust badges

UI Component Map

  • ui-component-map.md

Defines:

  • page layout
  • navigation structure
  • shared components
  • trace drawer structure
  • proposal/eval/replay modules

Acceptance Gates

  • acceptance-gates.md

Defines:

  • implementation gates
  • red flags
  • release conditions
  • mutation restrictions

Core Principles

Replay before persuasion

The UI should make replay evidence more important than aesthetic persuasion.

Calm by default

The interface should remain quiet and precise.

Progressive disclosure

The system should become more transparent the deeper an operator inspects.

Proposal is not ratification

The UI must preserve ADR-0057 review doctrine.

Read-only first

The first workbench phases prioritize visibility and auditability before any mutation capability.


Current Status

The planning package is merged on main via 404e694 (docs(workbench): CORE Workbench v1 planning architecture (ADR-0160)).

The prototype branch feat/w026-workbench-readonly-api is superseded and must not be used as the implementation base. It mixed W-026 with frontend and trace work, added auth and web-framework dependencies before the local read-only boundary was proven, and included placeholder replay/trace behavior that could be mistaken for evidence.

The next accepted implementation starts clean from main with W-026 only: dataclass schemas, repo-root-constrained readers, a standard-library local HTTP API, and route/read-model tests. W-027 and later phases build on that boundary after it is accepted.


W-026 Local Runbook

Start the read-only local API:

core workbench api

Verify liveness:

curl http://127.0.0.1:8765/health

Inspect each W-026 endpoint family:

curl http://127.0.0.1:8765/runtime/status
curl http://127.0.0.1:8765/artifacts
curl http://127.0.0.1:8765/artifacts/evals/contemplation_quality/contract.md
curl http://127.0.0.1:8765/proposals
curl http://127.0.0.1:8765/evals
curl http://127.0.0.1:8765/evals/contemplation_quality
curl -X POST http://127.0.0.1:8765/evals/run \
  -H 'Content-Type: application/json' \
  -d '{"lane":"contemplation_quality","version":"v1","split":"public"}'

Bind to a different local port:

core workbench api --port 9000

Non-local binds require an explicit operator flag:

core workbench api --host 0.0.0.0 --allow-nonlocal-bind

Initial Work Queue

Work Item Goal
W-026 Read-only API
W-027 Frontend shell
W-028 Chat + trace drawer
W-029 Proposal queue
W-030 Eval center
W-031 Replay theater

Guiding Question

The correct operator reaction to the workbench should be:

“Wait… I can actually inspect and replay the cognition.”

not:

“Cool chatbot.”


W-027 Frontend Shell Runbook

Start the full local workbench (API + frontend):

# Terminal 1 — API
uv run core workbench api               # http://127.0.0.1:8765

# Terminal 2 — Frontend
cd workbench-ui
pnpm install
pnpm dev                                # http://127.0.0.1:5173

Use a custom API URL:

VITE_WORKBENCH_API_URL=http://127.0.0.1:9000 pnpm dev

View the Branch 1 design baseline:

cd workbench-ui
pnpm preview   # Navigate to /preview

Detect TypeScript ↔ Python schema drift:

uv run python scripts/dump-api-schemas.py
cd workbench-ui && pnpm test