feat(framework): remove the architect (#545)

.changeset/dashboard-run-view.md

@@ -2,4 +2,4 @@
 "@gemstack/framework": minor
 ---
 
-Add event-stream projections for the dashboard's run overview: `architectPlan`, `decisionLedger`, `loopStatus`, and `sessionInfo` derive the chosen stack + rationale, the decisions ledger, the production-grade loop status, and the live session from a `FrameworkEvent[]`. They plus `formatFrameworkEvent` are re-exported from a new browser-safe `@gemstack/framework/client` subpath (no Node imports), so the dashboard renders the rich run view — the stack/PROS-CONS card, decisions, loop status, and a human-readable event log instead of raw JSON — across the live view, past-run replay, and the relay watch view.
+Add event-stream projections for the dashboard's run overview: `loopStatus` and `sessionInfo` derive the production-grade loop status and the live session from a `FrameworkEvent[]`. They plus `formatFrameworkEvent` are re-exported from a new browser-safe `@gemstack/framework/client` subpath (no Node imports), so the dashboard renders the rich run view (loop status and a human-readable event log instead of raw JSON) across the live view, past-run replay, and the relay watch view.

.changeset/interactive-choices-autopilot.md

@@ -2,6 +2,6 @@
 '@gemstack/framework': minor
 ---
 
-feat(framework): interactive plan-approval choice + autopilot in the dashboard (#304)
+feat(framework): interactive choice gate + autopilot in the dashboard (#304)
 
-The run now pauses at a plan-approval gate right after the architect decides the stack (the AWAIT point of the plan-then-AWAIT flow): the dashboard shows a "Your call" panel with "Proceed" plus each architect alternative as "Use X instead". Accept with the button or Ctrl+Enter, or leave the `[x] autopilot` countdown auto-accept the recommended plan after 10s (moving the mouse cancels it). Picking an alternative re-architects the run around it. New `requestChoice` option on `runFramework`, `choice` / `choice-resolved` events, and a `POST /choice` dashboard route; a headless run with no handler auto-accepts the recommended plan, so nothing else changes. Closes #304.
+A run can now pause on a choice and wait for you: the dashboard shows a "Your call" panel with the options. Accept with the button or Ctrl+Enter, or leave the `[x] autopilot` countdown auto-accept the recommended option after 10s (moving the mouse cancels it). New `requestChoice` option on `runFramework`, `choice` / `choice-resolved` events, and a `POST /choice` dashboard route; a headless run with no handler auto-accepts the recommended option, so nothing else changes. Closes #304.

.changeset/remove-architect-545.md

@@ -0,0 +1,16 @@
+---
+'@gemstack/framework': minor
+'@gemstack/ai-autopilot': minor
+---
+
+Remove the architect. A build no longer runs a turn to pick the app's stack, and no longer tells the agent what to build on: the agent reads the workspace and decides for itself, the same way it would outside The Framework. `buildPrompt` / `extendPrompt` / `scaffoldPrompt` now take just the intent and say nothing about a stack.
+
+Being opinionated about the stack is a hard thing to do well, and a system prompt that nudges one is worse than none at all (#545). The stack guidance we had was not designed, it accumulated. It goes rather than half-ships.
+
+Gone with it, because the architect was their only source:
+
+- the plan-approval gate ("Approve this plan?" / "Use X instead") and re-architect. The agent-authored await gates a build turn raises are unaffected: a build that stops to ask still pauses the run with Approve / Decline.
+- the decisions ledger in a run, and the `DECISIONS.md` a run wrote from it. `@gemstack/ai-autopilot`'s `decisions/` module is untouched and still exported; nothing in a run feeds it now.
+- the dashboard's "Stack & rationale" and "Decisions ledger" panels. Loop status, deploy, and session cards are unchanged.
+- `@gemstack/ai-autopilot`: `agentArchitect`, `STACK_TRADEOFFS`, `BootstrapSteps.architect`, `BootstrapOptions.ledger`, the `architect` bootstrap event, `ArchitectPlan` / `ArchitectContext` / `ArchitectDecision` / `ArchitectAlternative`, and the `plan` field on the build / loop / deploy contexts and on `BootstrapResult`. The flow is now scope -> build -> loop -> deploy.
+- `@gemstack/framework`: `driverArchitect`, `reArchitect`, `architectPrompt`, `parseArchitectPlan`, `architectPlan`, `decisionLedger`, `RunFrameworkResult.ledger`, and the `bench:architect` benchmark.

.changeset/research-preset-button.md

@@ -2,4 +2,4 @@
 '@gemstack/framework': minor
 ---
 
-The [Research] preset (#331): `framework research [what]` and a [Research] button on the daemon dashboard run Rom's problem-variability review as a direct prompt via the new `runPrompt()` path, which honors await gates (showChoices/showMultiSelect) but skips the scope/architect/build scaffolding. The "what" defaults to `this PR`.
+The [Research] preset (#331): `framework research [what]` and a [Research] button on the daemon dashboard run Rom's problem-variability review as a direct prompt via the new `runPrompt()` path, which honors await gates (showChoices/showMultiSelect) but skips the scope/build scaffolding. The "what" defaults to `this PR`.

.changeset/single-select-gate-primitive.md

@@ -4,4 +4,4 @@
 
 feat(framework): extract a shared single-select gate primitive (`requestChoices`) (#335)
 
-The single-select choice gate (#304) is now a reusable `requestChoices({ id, title, options, recommended })` export, the twin of `requestMultiSelect` (#332): it emits the `choice` event, parks for the pick, and falls back to the recommended option if the run is headless or aborts. The plan-approval gate builds on it, and it is the primitive the system prompt's `showChoices()` and the research preset need. No behavior change for existing runs.
+The single-select choice gate (#304) is now a reusable `requestChoices({ id, title, options, recommended })` export, the twin of `requestMultiSelect` (#332): it emits the `choice` event, parks for the pick, and falls back to the recommended option if the run is headless or aborts. It is the primitive the system prompt's `showChoices()` and the research preset need. No behavior change for existing runs.

.changeset/turn-boundary-choice-gate.md

@@ -4,4 +4,4 @@
 
 feat(framework): turn an agent's showChoices()+AWAIT into a live gate at the turn boundary (#337)
 
-The system prompt tells the agent to `showChoices()` and `AWAIT` at unclear-scope / alternatives points, but until now only framework-emitted gates (the plan-approval gate, multi-select) could pause a run. Now when a build turn ends by asking the user, an `await-choices` block, the framework shows the choice on the dashboard, waits for the pick, and re-prompts the agent to continue from that decision. It is the agent-authored counterpart to the plan-approval gate: a no-op when headless and when the agent just finishes instead of asking, so existing runs are unchanged.
+The system prompt tells the agent to `showChoices()` and `AWAIT` at unclear-scope / alternatives points, but until now only framework-emitted gates (multi-select) could pause a run. Now when a build turn ends by asking the user, an `await-choices` block, the framework shows the choice on the dashboard, waits for the pick, and re-prompts the agent to continue from that decision. It is a no-op when headless and when the agent just finishes instead of asking, so existing runs are unchanged.

Architecture.md

@@ -133,9 +133,8 @@ What shipped, built up from that seed (the state layer + the loop are the moat,
 - **Personas** — reusable, stack-aware roles materialized into worker agents; **presets** (Vike flagship, Next.js) select the framework-specific ones by detecting the project's framework, on top of a framework-neutral core.
 - **Runner** — a pluggable sandbox seam (`FakeRunner` + a real `LocalRunner`; Docker/WebContainer/Flue are infra-gated adapters) where agents build and run an app.
 - **Surfaces** — the same run in a terminal, an in-page UI, or a detached background handle, over one replayable event stream.
-- **Decisions ledger** — durable memory of rejected ideas + settled choices (round-trips `DECISIONS.md`) so a run stops re-pitching what was turned down.
 - **The loop** — an event-to-prompt-chain policy (a major change fires review + code-quality + security; a new UI flow fires QA + UX), gating on a `{ blockers }` verdict, with a data-driven **prompt library**.
-- **Bootstrap** — the spine that sequences the above into scope -> architect -> build -> full-fledged loop -> deploy, taking an app from nothing to production-grade.
+- **Bootstrap** — the spine that sequences the above into scope -> build -> full-fledged loop -> deploy, taking an app from nothing to production-grade.
 - **Scale mode** — a self-maintaining `CODE-OVERVIEW.md`, refreshed only on material change.
 - **Durability = in-process first.** Reuses `SubAgentRunStore` + resume primitives for pause/resume; a durable/queue-backed runner arrives as an optional adapter, not core.
 

docs/packages/ai-autopilot.md

@@ -17,7 +17,7 @@ The Supervisor (below) is the seed topology. Built up from it, the package is a
 - **Surfaces** — the same run in a terminal, an in-page UI, or a detached background handle, over one replayable event stream (`launchAutopilot`).
 - **Decisions ledger** — durable memory of rejected ideas + settled choices (round-trips `DECISIONS.md`) so a run stops re-pitching what was turned down.
 - **The loop** — an event-to-prompt-chain policy (a major change fires review + code-quality + security; a new UI flow fires QA + UX), gating on a `{ blockers }` verdict, with a data-driven built-in **prompt library**.
-- **Bootstrap** — the spine that sequences all of the above into scope → architect → build → full-fledged loop → deploy, taking an app from nothing to production-grade.
+- **Bootstrap** — the spine that sequences all of the above into scope → build → full-fledged loop → deploy, taking an app from nothing to production-grade.
 - **Scale mode** — a self-maintaining `CODE-OVERVIEW.md`, refreshed only on material change.
 
 See [`examples/bootstrap-quickstart`](https://github.com/gemstack-land/gemstack/tree/main/examples/bootstrap-quickstart) for all of it wired together offline.

examples/bootstrap-quickstart/README.md

@@ -5,7 +5,7 @@ project's dependencies to a scaffolded, production-grade, deploy-decided app.
 
 ```
 detect framework (preset)
-  → Bootstrap: scope → architect → build → full-fledged loop → deploy
+  → Bootstrap: scope → build → full-fledged loop → deploy
   → scale mode: CODE-OVERVIEW.md
 ```
 
@@ -26,12 +26,12 @@ ends at a live-looking URL with no credentials.
 
 - **Presets (#115)** — the framework is detected from the project deps (Vike here),
   so the build's workers are that framework's personas plus the shared neutral ones.
-- **Bootstrap (#116)** — one scoping question, then an **architect** picks the stack
-  and records its choices to the **decisions ledger**, a **build** scaffolds the app
+- **Bootstrap (#116)** — one scoping question, then a **build** scaffolds the app
   with the persona workers inside a **runner**, the **full-fledged loop** repeats the
   production-grade checklist until its `{ blockers }` verdict is empty, and a
   **deploy** is decided and shipped through the `DeployTarget` seam — here the real
   `cloudflareTarget` (SSR → Workers), run over a simulated `wrangler` offline.
+  What stack to build on is the build agent's call, not the framework's.
 - **Surfaces (#100/#120)** — every phase streams as narration over the generic
   `launchAutopilot<BootstrapEvent, BootstrapResult>` handle.
 - **Scale mode (#114)** — `CODE-OVERVIEW.md` is generated from the scaffold.
@@ -40,8 +40,8 @@ ends at a live-looking URL with no credentials.
 
 The offline run above verifies the flow structurally. `src/live.ts` runs the same
 flow for real — a real model via `@gemstack/ai-sdk` and a real `LocalRunner`
-workspace on the host filesystem, so the architect, the build workers, and the
-deploy decision are all model-driven and the workers write **real files to disk**:
+workspace on the host filesystem, so the build workers and the deploy decision are
+all model-driven and the workers write **real files to disk**:
 
 ```bash
 ANTHROPIC_API_KEY=sk-... pnpm --filter @gemstack/example-bootstrap-quickstart start:live

examples/bootstrap-quickstart/package.json

@@ -2,7 +2,7 @@
   "name": "@gemstack/example-bootstrap-quickstart",
   "version": "0.0.8",
   "private": true,
-  "description": "Runnable capstone for @gemstack/ai-autopilot: preset detection + Bootstrap (scope → architect → build → full-fledged loop → deploy) + scale mode, streamed over surfaces, offline via AiFake + FakeRunner.",
+  "description": "Runnable capstone for @gemstack/ai-autopilot: preset detection + Bootstrap (scope → build → full-fledged loop → deploy) + scale mode, streamed over surfaces, offline via AiFake + FakeRunner.",
   "type": "module",
   "scripts": {
     "typecheck": "tsc --noEmit",

examples/bootstrap-quickstart/src/bootstrap.test.ts

@@ -3,18 +3,14 @@ import assert from 'node:assert/strict'
 import { runCapstone, INTENT } from './bootstrap.js'
 
 describe('bootstrap capstone: the whole epic composes end-to-end (offline)', () => {
-  it('detects the preset, runs scope → architect → build → loop → deploy, and maps the code', async () => {
+  it('detects the preset, runs scope → build → loop → deploy, and maps the code', async () => {
     const lines: string[] = []
     const { detection, result, events, files, overview } = await runCapstone(line => lines.push(line))
 
     // Preset: the Vike framework was detected from the project deps.
     assert.equal(detection.preset?.name, 'vike')
     assert.equal(detection.framework, 'Vike')
 
-    // Architect: chose the stack and recorded its choices to the ledger.
-    assert.match(result.plan.stack, /Vike \+ Prisma/)
-    assert.equal(result.plan.decisions.length, 2)
-
     // Build: each preset persona wrote its file into the sandbox.
     assert.ok('database/schema.ts' in files)
     assert.ok('pages/orders/+Page.jsx' in files)