feat(pstack): give gh-stack inspect an ordered member list

README.md

@@ -14,7 +14,7 @@ O `pi-pstack` preserva o método verification-first do projeto original: context
 - `/poteto-mode` persistente durante a sessão ativa.
 - Feature maps, artifacts de navegador e receipts vinculados ao `HEAD` exato.
 - Evals cegos entre modelos, com hard assertions que o judge não pode ignorar.
-- Stacked PRs com `gh stack` por padrão e Graphite como backend opcional.
+- Stacked PRs com `gh stack`.
 - Benny em modo draft-only: ele pode preparar uma draft PR, mas nunca faz merge ou deploy.
 
 ## Requisitos
@@ -26,8 +26,7 @@ O `pi-pstack` preserva o método verification-first do projeto original: context
 - Bun para as ferramentas locais que o utilizam.
 - `portless` para fluxos locais que expõem serviços.
 - Playwright Chromium para verificação de navegador.
-- `github/gh-stack` como backend padrão de stacked PRs.
-- Graphite `gt` somente quando o backend opcional for selecionado.
+- `github/gh-stack` como backend de stacked PRs.
 - Um provider externo de Slack e tracker para executar Benny contra serviços reais.
 
 ## Instalação
@@ -84,7 +83,7 @@ O modo sticky vale somente para a sessão ativa. O Pi restaura o estado pelo his
 
 ## Stacked PRs
 
-`gh stack` é o backend padrão. O adapter Graphite só aparece quando `gt` está instalado. O pacote traduz operações para os CLIs oficiais e não mantém um segundo grafo de branches.
+`gh stack` é o backend de delivery. O pacote traduz operações para o CLI oficial e não mantém um segundo grafo de branches.
 
 O merge atômico exige um receipt por PR até o alvo. Gere cada receipt no checkout limpo do respectivo `HEAD` e mantenha os artifacts em caminhos imutáveis disponíveis durante a validação final. Se um digest de uma camada anterior não estiver disponível no checkout atual, o merge falha fechado.
 

extensions/poteto-mode.ts

@@ -51,8 +51,6 @@ export default function potetoModeExtension(pi: ExtensionAPI): void {
   }
 
   function restore(ctx: ExtensionContext): void {
-    // This closure is intentionally reset from the active branch on every
-    // session lifecycle event. It prevents state leaking across sessions.
     active = branchState(ctx);
   }
 
@@ -68,8 +66,6 @@ export default function potetoModeExtension(pi: ExtensionAPI): void {
 
   pi.on("session_start", async (_event, ctx) => {
     restore(ctx);
-    // Probe all host capability surfaces during startup without changing
-    // settings or failing print/JSON sessions. A task invocation fails closed.
     preflight(ctx);
   });
 
@@ -245,7 +241,7 @@ export default function potetoModeExtension(pi: ExtensionAPI): void {
     description:
       "Hash local verification, review, and eval artifacts into a receipt for the current repository HEAD.",
     parameters: Type.Object({
-      backend: Type.Union([Type.Literal("gh-stack"), Type.Literal("graphite")]),
+      backend: Type.Literal("gh-stack"),
       featureMapPath: Type.String({ minLength: 1 }),
       skillPath: Type.String({ minLength: 1 }),
       reviewPath: Type.String({ minLength: 1 }),
@@ -303,9 +299,9 @@ export default function potetoModeExtension(pi: ExtensionAPI): void {
     name: "pstack_delivery",
     label: "Run gated pstack delivery",
     description:
-      "Inspect a stack or run a receipt-gated gh-stack or Graphite mutation. Auto-merge also verifies the live PR head and checks.",
+      "Inspect a stack or run a receipt-gated gh-stack mutation. Auto-merge also verifies the live PR head and checks.",
     parameters: Type.Object({
-      backend: Type.Union([Type.Literal("gh-stack"), Type.Literal("graphite")]),
+      backend: Type.Literal("gh-stack"),
       operation: Type.Union([
         Type.Literal("inspect"),
         Type.Literal("prepare"),
@@ -344,8 +340,6 @@ export default function potetoModeExtension(pi: ExtensionAPI): void {
             paths.map((path) => loadEvidenceReceipt(path, ctx.cwd)),
           );
           if (params.operation === "auto-merge") {
-            if (params.backend !== "gh-stack")
-              return deliveryRejected("Graphite auto-merge lacks stack-wide receipt verification");
             if (!params.pullRequest)
               return deliveryRejected("auto-merge requires a pull request number");
             const stackResult = await pi.exec("gh", ["stack", "view", "--json"], {
@@ -368,7 +362,7 @@ export default function potetoModeExtension(pi: ExtensionAPI): void {
                 root: ctx.cwd,
                 repoIdentity: repository.stdout.trim(),
                 headSha: entry.headSha,
-                backend: params.backend as StackBackendName,
+                backend: params.backend,
                 level,
               });
               if (rejection) return deliveryRejected(rejection);
@@ -391,15 +385,15 @@ export default function potetoModeExtension(pi: ExtensionAPI): void {
               root: ctx.cwd,
               repoIdentity: repository.stdout.trim(),
               headSha: currentHead.stdout.trim(),
-              backend: params.backend as StackBackendName,
+              backend: params.backend,
               level,
             });
             if (rejection) return deliveryRejected(rejection);
             const authorization = authorizeDelivery({
               receipt,
               repoIdentity: repository.stdout.trim(),
               currentHeadSha: currentHead.stdout.trim(),
-              backend: params.backend as StackBackendName,
+              backend: params.backend,
               level,
             });
             if (authorization.draftOnly && params.operation === "submit" && params.draft === false)
@@ -414,18 +408,8 @@ export default function potetoModeExtension(pi: ExtensionAPI): void {
             return { exitCode: result.code, stdout: result.stdout, stderr: result.stderr };
           },
         };
-        const availableCommands = ["gh"];
-        if (params.backend === "graphite") {
-          const gt = await pi.exec("gt", ["--version"], { cwd: ctx.cwd, signal });
-          if (gt.code === 0) availableCommands.push("gt");
-        }
-        const backends = createDeliveryBackends({
-          runner,
-          availableCommands,
-        });
-        const backend = params.backend === "graphite" ? backends.graphite : backends.ghStack;
-        if (!backend)
-          return deliveryRejected("Graphite is unavailable; install and authenticate gt first");
+        const backends = createDeliveryBackends({ runner });
+        const backend = backends.ghStack;
         const operation = deliveryOperation(params);
         const result = await backend.execute(operation);
         return {
@@ -581,7 +565,6 @@ async function verifyPullRequestState(
   return undefined;
 }
 
-/** Flatten shell-like and structured tool payloads before applying merge gates. */
 function normalizeToolCallInput(input: unknown): string {
   const values: string[] = [];
   const seen = new WeakSet<object>();
@@ -692,6 +675,7 @@ function deliveryOperation(params: {
     case "rebase":
       return { kind: "rebase" };
     case "auto-merge":
+      if (!params.pullRequest) throw new Error("auto-merge requires pull request");
       return { kind: "auto-merge", pullRequest: params.pullRequest };
     default:
       throw new Error(`unsupported delivery operation: ${params.operation}`);

skills/poteto-mode/SKILL.md

@@ -17,9 +17,8 @@ managed worktrees isolate their edits. Resolve models through Pi profiles and
 agent roles, never through provider-specific slugs embedded in a skill.
 
 Delivery resources prepare evidence and receipts only. The default named
-backend is `gh stack`; Graphite is an optional named backend. Executable
-delivery behavior belongs to the later `src/delivery` adapters, not these
-skills.
+backend is `gh stack`. Executable delivery behavior belongs to the later
+`src/delivery` adapters, not these skills.
 
 ## Non-negotiables
 
@@ -139,7 +138,7 @@ A large or cross-cutting effort (a migration across many call sites, an ambitiou
 - **Authoring or modifying a skill.** Writing or editing a SKILL.md. `playbooks/authoring-a-skill.md`.
 - **Eval.** Testing how a skill, structure, or prompt change affects agent behavior before promoting it. `playbooks/eval.md`.
 - **Babysit.** Driving a PR or a stack to merge-ready: conflicts, review threads, CI. `playbooks/babysit.md`.
-- **Shipping.** The half after Babysit. Independently verifying a green stack, then handing the contiguous verified run to the default `gh stack` adapter (Graphite is optional). `playbooks/shipping.md`.
+- **Shipping.** The half after Babysit. Independently verifying a green stack, then handing the contiguous verified run to the `gh stack` adapter. `playbooks/shipping.md`.
 - **Autonomous run.** A long task to drive to completion without stopping ("run until done", "wait until X"). `playbooks/autonomous-run.md`.
 - **Orchestrate.** A standing project handed to one coordinator chat: multi-day, many stacked PRs, dozens to hundreds of subagents, minimal human turns ("run this whole project", "own this migration until it lands"). Distinct from Autonomous run, which drives one task to a predicate; work one agent could finish inside the session's budget routes there, not here, however program-shaped the phrasing sounds. `playbooks/orchestrate.md`.
 - **Autopilot-full.** A queue of independent PRs run to merged with full autonomy: one owner per PR carries build through merge, and the root swarm-verifies each merge-ready head before its owner merges ("autopilot this queue", "full autopilot", one-owner-per-PR programs). `playbooks/autopilot-full.md`.

skills/poteto-mode/playbooks/autopilot-stack.md

@@ -2,7 +2,7 @@
 
 **You own the stack, never the landing. Build and verify the queue with full
 autonomy, then hand the operator one linear reviewed stack.** `gh stack` is
-the default named backend; Graphite is optional. For "autopilot-stack", "stack
+the delivery backend. For "autopilot-stack", "stack
 them, don't ship", "build the stack, I'll land it". The sibling of
 **Autopilot-full**. The owner loop and verification gate are the same; only
 the terminal differs. Nothing auto-ships from this skill.

skills/poteto-mode/playbooks/opening-a-pr.md

@@ -20,7 +20,7 @@ Invoked at the end of every other playbook.
 
 After these sections, attach videos or screenshots when they prove a claim. Do not use `## Summary` or `## Test plan` boilerplate. A commit body does not restate its subject.
 
-**Size and stacks.** Prefer narrow PRs to one large PR. Use the configured delivery adapter, with `gh stack` as the default and Graphite as an explicit option. Keep the ordered stack visible to reviewers. Branch from main only for independent work. Rebase on `main` before substantial stack work.
+**Size and stacks.** Prefer narrow PRs to one large PR. Use the `gh stack` delivery adapter. Keep the ordered stack visible to reviewers. Branch from main only for independent work. Rebase on `main` before substantial stack work.
 
 **Readiness.** Open every PR ready, never as a draft. Set `draft: false` on every PR creation call. If a PR still opens as a draft, run the host's ready command, such as `gh pr ready <number>`. Run `gh pr view <number>` before you refer to PR status.
 

skills/poteto-mode/playbooks/shipping.md

@@ -4,17 +4,15 @@
 
 This is the half after `playbooks/babysit.md`. Babysit makes a stack mergeable.
 Shipping decides what is actually safe to merge and hands a verified receipt to
-the selected delivery adapter. `gh stack` is the default backend; Graphite is
-an optional named backend. These resources do not implement backend state.
+the selected delivery adapter. `gh stack` is the delivery backend. These resources do not implement backend state.
 Green is not safe, and the gap between those two words is where this playbook
 lives.
 
 1. **Verify every PR independently before arming anything.** One subagent per PR, not batched, each a managed-worktree subagent, each exercising the real surface through the available browser, CLI, desktop, or mobile verification skill against parent versus head. Each returns `PASS`, `PASS+NOTES` or `FAIL` and posts that verdict on its own PR so the record outlives the chat. Safe means a verdict from an agent that did not write the code. CI green is not a verdict, and an approving bot review is not a verdict.
 2. **Land only the contiguous verified run rooted at the bottom.** Walk up from the lowest unmerged PR and stop at the first one without a passing verdict, where both `PASS` and `PASS+NOTES` pass. A verified PR sitting above an unverified one is not landable, because merging it would pull the gap in underneath it. Report the ceiling as a PR number and say what breaks the chain.
 3. **Re-check that the verdicts still describe the code.** A restack rewrites every SHA above it and silently invalidates every verdict without touching a single check. Compare `git patch-id` at the verdict SHA against the current head before trusting an older verdict, and re-verify anything that actually drifted. Twenty-one verdicts went stale this way in one run with no signal at all.
-4. **Hand the receipt to the delivery adapter.** Use the default `gh stack`
-   adapter unless the operator explicitly selected the named Graphite adapter.
-   Do not invent or mutate stack state in a skill; the later
+4. **Hand the receipt to the delivery adapter.** Use the `gh stack`
+   adapter. Do not invent or mutate stack state in a skill; the later
    `src/delivery` adapter owns submission, merge-when-ready, and exact-head
    checks.
 5. **Never enable GitHub auto-merge on a stack directly.** Only the selected