carthagent 1.1.0-beta.2 → 1.1.0-beta.3

package/docs/beta2-published-acceptance.md

@@ -0,0 +1,79 @@
+# Beta.2 release acceptance
+
+November 2, 2026. Immutable tag `v1.1.0-beta.2` points to
+`e14aeb17c213a927a27e9522da65db5928219647`.
+
+## Completed
+
+- Native onboarding PR #238 and first-task guidance PR #239 merged. Release PR
+  #240 merged with tag ancestry preserved at `fc5382364f41b1563ed137ed0db0c5711e041a83`.
+- Five-platform prebuild [41505735348](https://github.com/B-icy/Carthagent/actions/runs/41505735348)
+  passed. Local manifest regeneration verified all archive/executable hashes and
+  embedded provenance against the pinned runtime recipe.
+- Release CI [41509102387](https://github.com/B-icy/Carthagent/actions/runs/41509102387)
+  passed across Linux, macOS and Windows, including installed-package acceptance.
+- Final draft acceptance [41511681097](https://github.com/B-icy/Carthagent/actions/runs/41511681097)
+  passed against the exact tag, assets and main ancestry.
+- GitHub prerelease is public with five runtime archives, the manifest and
+  combined checksums. Independent anonymous verification of all five public
+  archives passed.
+- Release workflow [41512736590](https://github.com/B-icy/Carthagent/actions/runs/41512736590)
+  passed anonymous installed-package acceptance and reports successful npm
+  publication with signed provenance (transparency log index 3648230788).
+- Backend onboarding #178 and retention reporting #179 are deployed. Live
+  authenticated account identity and read-only plan terms were verified earlier.
+
+## npm publication verified
+
+After the workflow reported `+ carthagent@1.1.0-beta.2`, public registry requests
+from the acceptance environment still returned 404 for that exact version.
+Both npm CLI and cache-bypassed package metadata initially reported
+`next=1.1.0-beta.1`, `latest=1.0.7`. Registry visibility subsequently converged;
+no repeat publication was attempted.
+
+The public beta.2 tarball's SHA-512 matches registry integrity and the provenance
+subject digest. Sigstore verification passed for the GitHub Actions issuer and
+the repository's `release.yml@refs/heads/main` identity. The statement identifies
+release workflow 41512736590, and package `gitHead` matches the immutable tag.
+Verified tags: `next=1.1.0-beta.2`, `latest=1.0.7`.
+Evidence: `/tmp/opencode/beta2-npm-acceptance.json`.
+
+## Installed update verified
+
+The website command `npm install -g carthagent@next` installed beta.2 into an
+isolated Linux x64 home/prefix. Default runtime acquisition resolved the pinned
+CTG-owned executable with SHA-256
+`8a34735274c0b62049f6ae882dc955016fde54458efc18822fbcedb2e8eddeda`.
+The beta update check reported the installed version as current.
+Evidence: `/tmp/opencode/beta2-clean-install-acceptance.json`.
+
+An isolated npm installation of beta.1 detected beta.2 with
+`ctg update --channel beta --check`. Running `ctg update --channel beta` upgraded
+it to beta.2; a profile sentinel remained byte-for-byte intact. This used a
+disposable home and global npm prefix, not the user's normal installation.
+Evidence: `/tmp/opencode/beta2-update-acceptance.json`.
+
+## Remaining acceptance
+
+Stripe sandbox access subsequently became available. Real purchases, upgrades,
+SCA, renewal, failed renewal and cancellation were exercised separately against
+the isolated backend (see backend `docs/stripe-sandbox-acceptance.md`).
+
+An isolated published beta.2 login connected the authorized existing Cloud account.
+A real range-fix task passed four independent tests. Continuing that same session
+hit its persistent elapsed-time budget; beta.2 showed only an unexplained interrupt.
+The follow-up source fix adds actionable budget diagnostics and explicit
+`--reset-budget` support. With that source launcher and the same installed runtime,
+profile and session, continuation added an ascending-across-zero regression and
+passed all five independent tests. This launcher fix is not part of immutable beta.2.
+Execution budgets are not billing credit and never reset on payment or ordinary
+continuation. Evidence: `/tmp/opencode/stripe-acceptance/live-continuation-fixed.log`.
+
+Full task continuation across a genuine payment boundary remains incomplete:
+the real model task used existing Cloud credit and sandbox payments used a separate
+backend without a model upstream. Screenshot approval is not claimed. Local terminal timing failures
+were followed by passing full local, CI and draft acceptance; this does not prove
+that every terminal timing issue is resolved.
+
+The normal installed profile and Invest-With-Friends files were not changed by
+these release checks. Stable npm `latest` remains 1.0.7.

package/docs/cloud-onboarding-acceptance.md

@@ -1,7 +1,10 @@
-# Cloud onboarding — unreleased implementation
+# Cloud onboarding — beta.2 implementation
 
-November 2, 2026. This branch is not published beta acceptance. npm `next` remains
-1.1.0-beta.1 and `latest` remains 1.0.7.
+November 2, 2026. Beta.2 is published on GitHub and npm `next`; `latest` remains
+1.0.7. See [published acceptance](beta2-published-acceptance.md) for final CI,
+draft/public runtime gates, signed npm provenance, Linux clean installation and
+isolated beta.1 → beta.2 update evidence. Historical preparation notes below do
+not supersede that release record.
 
 Native PR #238 merged at `a5ad2164d056fee7f7183579ce6cc97305bae53b` after CI
 41504674623 passed. First-task guidance is tracked in PR #239. Beta.2 runtime

package/docs/release-notes-beta.3.md

@@ -0,0 +1,47 @@
+# Carthagent 1.1.0-beta.3 — release candidate
+
+This candidate makes continuation failures actionable. Publication is pending;
+these notes do not establish npm availability.
+
+## Continue the existing task
+
+- An exhausted persistent execution budget now reports the reason and recovery
+  command instead of an unexplained interrupt.
+- `--reset-budget` explicitly authorizes a fresh budget for a resumed session:
+
+  ```sh
+  ctg --print --continue --reset-budget "Continue the task"
+  ```
+
+- Ordinary continuation and payment preserve the execution budget, including
+  elapsed idle time. Billing credit is separate. Interactive users can run
+  `/delivery-budget-reset` before continuing the existing task.
+- CLI help consistently names Carthagent Cloud and documents `billing refresh`.
+
+## Acceptance evidence
+
+The source fix passed 1,132 tests with five skips, builds and lint. PR #242 CI
+passed Linux, macOS and Windows, browser and installed-package gates. One macOS
+22 service-stop identity timeout passed on retry; the separate push run passed.
+
+The authorized existing Cloud account completed a real range-fix task. The same
+session, profile and installed beta.2 runtime then continued with the source
+launcher and explicit reset, adding a regression and passing all five project
+tests independently. Final beta.3 installed acceptance remains required.
+
+Backend PRs #180–181 are deployed: automatic payment-status checks, clearer
+sign-in and refund guidance, direct Builder review, modern Stripe compatibility
+and accurate scheduled-cancellation dates while retaining paid access.
+Sandbox purchases, prorated upgrades, initial and upgrade SCA, declined-upgrade
+recovery, renewal, failed renewal, automatic refunds and disputes were exercised
+separately. These do not establish a single
+useful-task → payment boundary → same-task continuation journey: the model task
+used existing production credit and payments used an isolated sandbox backend.
+
+## Release channel
+
+Targets npm `next`; stable `latest` remains 1.0.7. Runtime prebuild
+[41537249365](https://github.com/B-icy/Carthagent/actions/runs/41537249365) uses
+reviewed source `c11103f8b57c6b3254473a78d535a404bea9a84a` and the unchanged
+combined-v5 runtime recipe. Final assets and manifest must pass immutable draft
+acceptance before publication.

package/lib/opencode/cli-runtime.mjs

@@ -100,6 +100,7 @@ function errorText(error) {
 function headlessFailureMessage(latest, headlessError) {
   const raw = errorText(latest?.error ?? latest?.session?.error ?? lastAssistantError(latest?.messages) ?? headlessError);
   const outcome = latest?.outcome ?? latest?.session?.outcome;
+  if (['elapsed-time', 'tool-calls', 'repair-rounds'].includes(latest?.budgetReason)) return `Session execution budget exhausted: ${latest.budgetReason}. Billing refresh does not reset execution limits. To authorize a fresh budget, resume this session with --reset-budget (for example: ctg --print --continue --reset-budget "Continue the task"), or run /delivery-budget-reset in the existing terminal session before continuing.`;
   if (/ModelNotSelectedError|No model is available|no provider|provider.*not (connected|selected|configured)|not authenticated/i.test(raw ?? '')) {
     return 'No AI provider connected. Run `ctg login` to connect one (Carthagent Cloud is recommended), or set an API key env var (e.g. ANTHROPIC_API_KEY), then retry. Run `ctg native --list-models` to see available models.';
   }
@@ -559,6 +560,9 @@ export async function createOpenCodeCliRuntime({ cwd = process.cwd(), io = conso
           options.session = await latestSessionForWorkspace(client, cwd);
           if (!options.session) throw new Error('No native Carthagent session exists for this workspace to continue.');
         }
+        if (options.resetBudget && !options.session && !options.continue && !options.resume) {
+          throw new Error('--reset-budget requires --session, --continue or --resume.');
+        }
         eventController = new AbortController();
         const nativeEvents = [];
         let eventFailure;
@@ -600,6 +604,9 @@ export async function createOpenCodeCliRuntime({ cwd = process.cwd(), io = conso
           const definitions = native.rpcDefinitions ?? await nativeRpcDefinitions();
           await client.rpc(definitions.SessionControls).framing({ sessionID, enabled: options.autoFraming }, { location: { directory: cwd }, signal: signal.controller.signal });
         }
+        if (options.resetBudget && sessionID) {
+          await client.session.command({ sessionID, name: 'delivery-budget-reset', text: '' }, { location: { directory: cwd }, signal: signal.controller.signal });
+        }
         if (framedPrompt) {
           if (!sessionID) {
             const created = await headlessApi.createSession(service, cwd, { body: {
@@ -644,6 +651,11 @@ export async function createOpenCodeCliRuntime({ cwd = process.cwd(), io = conso
         const definitions = native.rpcDefinitions ?? await nativeRpcDefinitions();
         await client.rpc(definitions.SessionControls).framing({ sessionID: requestedSession, enabled: options.autoFraming }, { location: { directory: cwd }, signal: signal.controller.signal });
       }
+      if (requestedSession && options.resetBudget && !signal.controller.signal.aborted) {
+        const resumed = await client.session.get({ sessionID: requestedSession }, { signal: signal.controller.signal });
+        assertSessionWorkspace(resumed, cwd);
+        await client.session.command({ sessionID: requestedSession, name: 'delivery-budget-reset', text: '' }, { location: { directory: cwd }, signal: signal.controller.signal });
+      }
       try {
         result = await headlessApi.runHeadless(service, cwd, framedPrompt || prompt || '', {
           sessionID: options.session,
@@ -658,7 +670,16 @@ export async function createOpenCodeCliRuntime({ cwd = process.cwd(), io = conso
         headlessError = error;
         result = { session: { id: options.session }, messages: [], outcome: signal.controller.signal.aborted ? 'interrupted' : 'failed', terminal: true, error: error.message };
       }
-        const latest = headlessError ? result : await reconcileHeadlessResult(result, { client, signal: signal.controller.signal });
+        let latest = headlessError ? result : await reconcileHeadlessResult(result, { client, signal: signal.controller.signal });
+        if (latest.outcome === 'interrupted' && !latest.error && !latest.session?.error && !lastAssistantError(latest.messages) && latest.session?.id && typeof client.rpc === 'function' && !signal.controller.signal.aborted) {
+          // An interrupt may carry no error in the host projection. Read the
+          // authoritative persistent budget without restarting the task.
+          try {
+            const definitions = native.rpcDefinitions ?? await nativeRpcDefinitions();
+            const budget = await client.rpc(definitions.Carthagent).budgetStatus({ sessionID: latest.session.id }, { location: { directory: cwd }, signal: AbortSignal.any([signal.controller.signal, AbortSignal.timeout(2000)]) });
+            if (budget?.stopped) latest = { ...latest, budgetReason: budget.stopped, error: latest.error ?? headlessFailureMessage({ budgetReason: budget.stopped }) };
+          } catch { /* Keep the original failure when budget status is unavailable. */ }
+        }
         eventController.abort();
         await eventReader;
         if (eventFailure) throw new Error(`Native Carthagent event stream failed: ${eventFailure.message}`, { cause: eventFailure });

package/lib/opencode/cli.mjs

@@ -26,6 +26,7 @@ const BOOL_FLAGS = new Map([
   ['--protect-existing', ['protectExisting', true]], ['--jev', ['jev', true]], ['--no-jev', ['jev', false]],
   ['--print', ['print', true]], ['--demo', ['demo', true]], ['-c', ['continue', true]],
   ['--continue', ['continue', true]], ['-r', ['resume', true]], ['--resume', ['resume', true]],
+  ['--reset-budget', ['resetBudget', true]],
 ]);
 
 const aliases = new Map([
@@ -84,9 +85,9 @@ export function cliHelp(version = '') {
   \x1b[32mhash\x1b[0m, \x1b[32mfingerprint\x1b[0m     Calculate workspace SHA-256 fingerprint
   \x1b[32mvalidate\x1b[0m <file.json> Validate acceptance contract
   \x1b[32mlogin\x1b[0m, \x1b[32mauth\x1b[0m             Connect an AI provider
-  \x1b[32maccount\x1b[0m               Show Ship credit, plan, grants and CLI sessions
-  \x1b[32mbilling\x1b[0m [checkout|portal] Open Ship billing
-  \x1b[32mlogout-cloud\x1b[0m [session] Revoke the current or named Ship session
+  \x1b[32maccount\x1b[0m               Show Cloud credit, plan, grants and CLI sessions
+  \x1b[32mbilling\x1b[0m [checkout|portal|refresh] Open or refresh Cloud billing
+  \x1b[32mlogout-cloud\x1b[0m [session] Revoke the current or named Cloud session
   \x1b[32marchive\x1b[0m [list|search|read|export] Read-only legacy session archive
   \x1b[32mreview\x1b[0m                Configure self-review or review a PR
   \x1b[32mserve\x1b[0m, \x1b[32mweb\x1b[0m, \x1b[32mstart\x1b[0m     Start the web dashboard
@@ -110,6 +111,8 @@ export function cliHelp(version = '') {
   --max-tools <n>          Session tool-call budget (0 disables)
   --max-seconds <n>        Session elapsed-time budget (0 disables)
   --max-repairs <n>        Automatic-repair budget (0 disables)
+  --reset-budget           Explicitly reset a resumed session's execution budget
+                           Budgets persist across continuation, including idle time.
   --protect-existing      Preserve files present at session start
   --rewrite-cap <n>        Whole-file writes per path (0 disables)
   --tool-output-cap <n>    Bound text output per tool result

package/lib/opencode/runtime-release.json

@@ -1,6 +1,6 @@
 {
   "version": 1,
-  "releaseVersion": "1.1.0-beta.2",
+  "releaseVersion": "1.1.0-beta.3",
   "compatibilityVersion": "2.0.32",
   "runtimeVersion": "0.0.0-ctg.2.0.32",
   "sourceRevision": "843a6c514e606b436a7fe713422ace0a93ad3d61",
@@ -9,28 +9,28 @@
   "assets": {
     "linux-x64": {
       "name": "opencode-linux-x64.tar.gz",
-      "archiveSha256": "0389612a88f9b002b819e4fc12eb5bd9e337b9c5352de60dd3f523e343ac01c2",
-      "executableSha256": "8a34735274c0b62049f6ae882dc955016fde54458efc18822fbcedb2e8eddeda"
+      "archiveSha256": "ab95bd99fc2bf52f867927a817dde1d3882122b80f85ef85cf416733afa62483",
+      "executableSha256": "f9ad13d8aad76bf1d16334712f0060b886757f5cdabddacce42f9805d4482c1c"
     },
     "linux-arm64": {
       "name": "opencode-linux-arm64.tar.gz",
-      "archiveSha256": "7a01012e830a2bee2195402bf0c525bd345754ee4abb72af3ba4a597ceb096e5",
-      "executableSha256": "aea44aa7d12ee9aef5e6fd21fbfe1d23bd2aa2c731d955d5393c4fe2c4b1dcd7"
+      "archiveSha256": "3869dd9398693afffe9f438192fbb12061b66a211993018379fb6870d2012bd2",
+      "executableSha256": "0c302d55bba608d83c42f54b769ec3994421d3dcb16ea5d2a9455dcd91f884a1"
     },
     "darwin-x64": {
       "name": "opencode-darwin-x64.tar.gz",
-      "archiveSha256": "cf88ba7c06a75a9e2021cfbb783a23fb93f09e2fb2dce86067346c8c666a2c5b",
-      "executableSha256": "17c67ad68d84a9c489a4d3567784aa8a2e3f7b086153693a57c1ef0a8b270a4f"
+      "archiveSha256": "b20111d8fd2368ac9bde7f90b12835a465b8c92da3db1f3a35671860a325eb27",
+      "executableSha256": "23f7a845e75592efa11dff407e147287584eae7f63fca1a9c4ccbe9558656bc3"
     },
     "darwin-arm64": {
       "name": "opencode-darwin-arm64.tar.gz",
-      "archiveSha256": "8db89fcec1f4fe502dfbb6f0e9ad6e2ccecdc6191ad14d6afa5338e9e0990341",
-      "executableSha256": "789dc613d191d339ef4b789d36dd32403db9422f11841659aea5a9ec4d3e6e37"
+      "archiveSha256": "eb65c635c8cf19f0830b56b2d8087387148edaf44cf821c881b72388c9c185fd",
+      "executableSha256": "d10a7b549748a131c578ed8cd041aeb652f1ae9c79c5f3912a731f2a69218736"
     },
     "windows-x64": {
       "name": "opencode-windows-x64.tar.gz",
-      "archiveSha256": "9f7139dea76e662109a4c80b109b92525c993a37c69aae5b45623f8d69ab9467",
-      "executableSha256": "38edbca2f4304f2efa2afa57b654c362c14f422c141c3ebd82e1062221dccd80"
+      "archiveSha256": "a0309761781f57b5ea9f329d41f2e977391949addc9baaca6b85025780249d3c",
+      "executableSha256": "91eac290633313c7fce4a426ef265f193e995b285746df6a2a6010f4727fe813"
     }
   }
 }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "carthagent",
-  "version": "1.1.0-beta.2",
+  "version": "1.1.0-beta.3",
   "workspaces": [
     "plugins/carthagent"
   ],

package/tests/cli.test.mjs

@@ -102,11 +102,11 @@ test('check fails clearly when no delivery report exists', t => {
   assert.match(result.stderr, /No delivery report/);
 });
 
-test('guidance routing is not a user-facing command and Ship account commands are documented', t => {
+test('guidance routing is not a user-facing command and Cloud account commands are documented', t => {
   const result = run(fixture(t), '--help');
   assert.equal(result.status, 0, result.stderr);
   assert.doesNotMatch(result.stdout, /\broute\b|\bprofiles\b/);
-  assert.match(result.stdout, /account.*Ship credit/);
+  assert.match(result.stdout, /account.*Cloud credit/);
   assert.match(result.stdout, /billing.*checkout\|portal/);
   assert.match(result.stdout, /logout-cloud/);
 });

package/tests/opencode-cli.test.mjs

@@ -42,6 +42,40 @@ test('native CLI selects headless execution and preserves output-mode options',
   assert.equal(invocation.prompt, 'Do work');
 });
 
+test('headless continuation explains persistent budget exhaustion and resets only on explicit request', async t => {
+  const { mkdtempSync, rmSync } = await import('node:fs');
+  const { tmpdir } = await import('node:os');
+  const cwd = mkdtempSync(join(tmpdir(), 'ctg-budget-continuation-'));
+  t.after(() => rmSync(cwd, { recursive: true, force: true }));
+  const { io, output } = ioCapture();
+  const calls = [];
+  let stopped = 'elapsed-time';
+  const runtime = await createOpenCodeCliRuntime({ cwd, io, stdout: { write() {} }, native: {
+    runtime: {}, service: { ensureService: async () => ({}) }, rpcDefinitions: { Carthagent: {} },
+    clientFactory: async () => ({
+      rpc: () => ({ budgetStatus: async () => ({ stopped }) }),
+      session: {
+        list: async () => ({ data: [{ id: 'same-session', location: { directory: cwd } }] }),
+        get: async () => ({ id: 'same-session', location: { directory: cwd } }),
+        command: async input => { calls.push(input); stopped = null; },
+      },
+    }),
+    headless: { runHeadless: async (_service, _cwd, _prompt, options) => {
+      assert.equal(options.sessionID, 'same-session');
+      calls.push('prompt');
+      return { session: { id: 'same-session' }, outcome: stopped ? 'interrupted' : 'succeeded', terminal: true,
+        messages: stopped ? [] : [{ type: 'assistant', finish: 'stop', content: [{ type: 'text', text: 'continued' }] }] };
+    } },
+  } });
+  assert.equal(parseOpenCodeArgs(['--continue', '--reset-budget', 'Continue']).resetBudget, true);
+  assert.equal(await runtime.run({ mode: 'headless', options: { continue: true }, prompt: 'Continue' }), 1);
+  assert.match(output.stderr.at(-1), /budget exhausted: elapsed-time.*--reset-budget/);
+  assert.deepEqual(calls, ['prompt'], 'no implicit budget reset or prompt retry');
+  assert.equal(await runtime.run({ mode: 'headless', options: { continue: true, resetBudget: true }, prompt: 'Continue' }), 0);
+  assert.deepEqual(calls.slice(1), [{ sessionID: 'same-session', name: 'delivery-budget-reset', text: '' }, 'prompt']);
+  await assert.rejects(runtime.run({ mode: 'headless', options: { resetBudget: true }, prompt: 'New task' }), /requires --session/);
+});
+
 test('headless CLI validates prompts and maps cancellation to conventional exit code', async () => {
   const { io, output } = ioCapture();
   assert.equal(await runOpenCodeCli({ argv: [], stdoutIsTTY: false, io, runtime: {} }), 1);
@@ -77,10 +111,10 @@ test('model, auth, cloud and domain entrypoints dispatch to injected native capa
   ]);
 });
 
-test('help preserves command aliases, Ship operations and isolation semantics', async () => {
+test('help preserves command aliases, Cloud operations and isolation semantics', async () => {
   const { output, io } = ioCapture();
   assert.equal(await runOpenCodeCli({ argv: ['--help'], io, version: '1.0.3' }), 0);
-  assert.match(output.stdout[0], /account.*Ship credit/);
+  assert.match(output.stdout[0], /account.*Cloud credit/);
   assert.match(output.stdout[0], /billing.*checkout\|portal/);
   assert.match(output.stdout[0], /--isolate\s+Isolate extension\/skill discovery \(not a worktree\)/);
   assert.equal(cliHelp('x').includes('--engine-cli, --agent-cli'), true);