@fitlab-ai/agent-infra 0.9.12 → 0.9.13

package/README.md

@@ -55,10 +55,10 @@ Once initialized, open the project in your AI TUI and install the latest skills:
 
 ```bash
 /import-issue 42           # AI reads the issue, creates a task, extracts requirements
-/analyze-task <task-id>    # AI scans the codebase, finds the root cause, writes analysis.md
-/review-analysis <task-id> # An isolated reviewer checks the analysis
-/plan-task <task-id>       # AI proposes a fix plan
-/review-plan <task-id>     # A fresh isolated reviewer checks the plan
+/analyze-task --task <task-id>    # AI scans the codebase, finds the root cause, writes analysis.md
+/review-analysis --task <task-id> # An isolated reviewer checks the analysis
+/plan-task --task <task-id>       # AI proposes a fix plan
+/review-plan --task <task-id>     # A fresh isolated reviewer checks the plan
 ```
 
 > **You review the plan and reply in natural language:**
@@ -71,12 +71,12 @@ Just fix it at the application layer in LoginService.
 > AI re-runs `/plan-task` to update the plan accordingly and confirms.
 
 ```bash
-/code-task <task-id>       # AI writes the fix, tests it, and creates a local checkpoint commit
-/review-code <task-id>     # A fresh isolated reviewer reports one minor finding
-/code-task <task-id>       # AI fixes the minor issue, re-validates, and creates a new checkpoint
-/review-code <task-id>     # The reviewer approves the current checkpoint
-/create-pr <task-id>       # The reviewed branch is published to its task-bound target; PR links issue #42
-/complete-task <task-id>   # task archived
+/code-task --task <task-id>       # AI writes the fix, tests it, and creates a local checkpoint commit
+/review-code --task <task-id>     # A fresh isolated reviewer reports one minor finding
+/code-task --task <task-id>       # AI fixes the minor issue, re-validates, and creates a new checkpoint
+/review-code --task <task-id>     # The reviewer approves the current checkpoint
+/create-pr --task <task-id>       # The reviewed branch is published to its task-bound target; PR links issue #42
+/complete-task --task <task-id>   # task archived
 ```
 
 **11 commands. 1 natural-language correction. From issue to merged PR.** That is the entire SOP — programming can have a standard operating procedure too.
@@ -192,7 +192,7 @@ The most-used lifecycle commands, in delivery order. The command prefix varies b
 | `code-task` → `review-code` | Implement and test; `code-task` creates a local checkpoint, then a structured review checks that checkpoint |
 | `create-pr` → `complete-task` | Publish the approved checkpoint to the task-bound target branch, then archive after merge and final gates |
 
-Start with one atomic role policy, for example: `$run-task 42 --executor-model <id> --executor-reasoning-effort <value> --reviewer-model <id> --reviewer-reasoning-effort <value>`. Both roles may use the same model. With no explicit policy, `run-task` reads the current Agent Client's optional `agentClients[].orchestration`; if neither source is complete, it shows the host model-selection guidance before creating any run. Re-entry uses the persisted run policy. Finish or clear active runs before upgrading agent-infra; unsupported disk state fails closed without rewriting.
+Start with one atomic role policy, for example: `$run-task --task 42 --executor-model <id> --executor-reasoning-effort <value> --reviewer-model <id> --reviewer-reasoning-effort <value>`. Both roles may use the same model. With no explicit policy, `run-task` reads the current Agent Client's optional `agentClients[].orchestration`; if neither source is complete, it shows the host model-selection guidance before creating any run. Re-entry uses the persisted run policy. Finish or clear active runs before upgrading agent-infra; unsupported disk state fails closed without rewriting.
 
 See the full catalog — task status, release, security, and project-maintenance skills — in [Built-in AI Skills](./docs/en/skills.md).
 

package/README.zh-CN.md

@@ -54,10 +54,10 @@ agent-infra 的目标就是把这层共享基础设施标准化。它为所有
 
 ```bash
 /import-issue 42           # AI 读取 Issue，创建任务，提取需求
-/analyze-task <task-id>    # AI 扫描代码库，定位根因，输出 analysis.md
-/review-analysis <task-id> # 隔离 reviewer 审查分析产物
-/plan-task <task-id>       # AI 提出修复方案
-/review-plan <task-id>     # fresh 隔离 reviewer 审查方案
+/analyze-task --task <task-id>    # AI 扫描代码库，定位根因，输出 analysis.md
+/review-analysis --task <task-id> # 隔离 reviewer 审查分析产物
+/plan-task --task <task-id>       # AI 提出修复方案
+/review-plan --task <task-id>     # fresh 隔离 reviewer 审查方案
 ```
 
 > **你审查方案后用自然语言回复：**
@@ -70,12 +70,12 @@ agent-infra 的目标就是把这层共享基础设施标准化。它为所有
 > AI 按你的要求重跑 `/plan-task` 更新方案并确认。
 
 ```bash
-/code-task <task-id>       # AI 编写修复代码、测试，并创建本地 checkpoint commit
-/review-code <task-id>     # fresh 隔离 reviewer 报告一个次要问题
-/code-task <task-id>       # AI 修复次要问题、重新验证，并创建新的 checkpoint
-/review-code <task-id>     # reviewer 审查并通过当前 checkpoint
-/create-pr <task-id>       # 将已审查分支发布到任务绑定目标分支；PR 自动关联 Issue #42
-/complete-task <task-id>   # 任务归档
+/code-task --task <task-id>       # AI 编写修复代码、测试，并创建本地 checkpoint commit
+/review-code --task <task-id>     # fresh 隔离 reviewer 报告一个次要问题
+/code-task --task <task-id>       # AI 修复次要问题、重新验证，并创建新的 checkpoint
+/review-code --task <task-id>     # reviewer 审查并通过当前 checkpoint
+/create-pr --task <task-id>       # 将已审查分支发布到任务绑定目标分支；PR 自动关联 Issue #42
+/complete-task --task <task-id>   # 任务归档
 ```
 
 **11 条命令，1 次自然语言纠正，从 Issue 到合并 PR。** 这就是完整的 SOP —— 编程也可以有标准作业流程。
@@ -191,7 +191,7 @@ ai agent-client configure
 | `code-task` → `review-code` | 实现并测试；`code-task` 创建本地 checkpoint，再由结构化审查检视该 checkpoint |
 | `create-pr` → `complete-task` | 将已通过审查的 checkpoint 发布到任务绑定目标分支，并在合并和最终门禁通过后归档 |
 
-启动时可提供原子角色策略，例如：`$run-task 42 --executor-model <id> --executor-reasoning-effort <值> --reviewer-model <id> --reviewer-reasoning-effort <值>`。两个角色可以使用同一模型。完全没有显式策略时，`run-task` 读取当前 Agent Client 可选的 `agentClients[].orchestration`；两个来源都不完整时，会在创建任何 run 前展示宿主模型选择指引。重入使用已持久化的 run 策略。升级 agent-infra 前必须完成或清空 active run；不受支持的磁盘状态会失败关闭且不改写。
+启动时可提供原子角色策略，例如：`$run-task --task 42 --executor-model <id> --executor-reasoning-effort <值> --reviewer-model <id> --reviewer-reasoning-effort <值>`。两个角色可以使用同一模型。完全没有显式策略时，`run-task` 读取当前 Agent Client 可选的 `agentClients[].orchestration`；两个来源都不完整时，会在创建任何 run 前展示宿主模型选择指引。重入使用已持久化的 run 策略。升级 agent-infra 前必须完成或清空 active run；不受支持的磁盘状态会失败关闭且不改写。
 
 完整清单（任务状态、发布、安全、项目维护等 skill）见 [内置 AI Skills](./docs/zh-CN/skills.md)。
 

package/dist/bin/internal-cli.js

@@ -45,12 +45,12 @@ if (taskControlCommand) {
             }
             case 'task-lifecycle': {
                 const { taskLifecycle } = await import('../lib/internal/task-lifecycle.js');
-                taskLifecycle(process.argv.slice(3));
+                await taskLifecycle(process.argv.slice(3));
                 break;
             }
             case 'task-finalization': {
                 const { taskFinalization } = await import('../lib/internal/task-finalization.js');
-                taskFinalization(process.argv.slice(3));
+                await taskFinalization(process.argv.slice(3));
                 break;
             }
         }
@@ -60,7 +60,7 @@ else
     switch (command) {
         case 'task-create': {
             const { taskCreate } = await import('../lib/internal/task-create.js');
-            taskCreate(process.argv.slice(3));
+            await taskCreate(process.argv.slice(3));
             break;
         }
         case 'sandbox-control': {
@@ -100,37 +100,37 @@ else
         }
         case 'platform-release-notes': {
             const { platformReleaseNotes } = await import('../lib/internal/platform-release-notes.js');
-            platformReleaseNotes(process.argv.slice(3));
+            await platformReleaseNotes(process.argv.slice(3));
             break;
         }
         case 'platform-context': {
             const { platformContext } = await import('../lib/internal/platform-context.js');
-            platformContext(process.argv.slice(3));
+            await platformContext(process.argv.slice(3));
             break;
         }
         case 'platform-comment': {
             const { platformComment } = await import('../lib/internal/platform-comment.js');
-            platformComment(process.argv.slice(3));
+            await platformComment(process.argv.slice(3));
             break;
         }
         case 'platform-issue': {
             const { platformIssue } = await import('../lib/internal/platform-issue.js');
-            platformIssue(process.argv.slice(3));
+            await platformIssue(process.argv.slice(3));
             break;
         }
         case 'platform-pr': {
             const { platformPr } = await import('../lib/internal/platform-pr.js');
-            platformPr(process.argv.slice(3));
+            await platformPr(process.argv.slice(3));
             break;
         }
         case 'platform-pr-review': {
             const { platformPrReview } = await import('../lib/internal/platform-pr-review.js');
-            platformPrReview(process.argv.slice(3));
+            await platformPrReview(process.argv.slice(3));
             break;
         }
         case 'pr-review-grade': {
             const { prReviewGrade } = await import('../lib/internal/pr-review-grade.js');
-            prReviewGrade(process.argv.slice(3));
+            await prReviewGrade(process.argv.slice(3));
             break;
         }
         case 'platform-checks': {
@@ -145,17 +145,17 @@ else
         }
         case 'task-ledger': {
             const { taskLedger } = await import('../lib/internal/task-ledger.js');
-            taskLedger(process.argv.slice(3));
+            await taskLedger(process.argv.slice(3));
             break;
         }
         case 'task-warning': {
             const { taskWarning } = await import('../lib/internal/task-warning.js');
-            taskWarning(process.argv.slice(3));
+            await taskWarning(process.argv.slice(3));
             break;
         }
         case 'task-activity': {
             const { taskActivity } = await import('../lib/internal/task-activity.js');
-            taskActivity(process.argv.slice(3));
+            await taskActivity(process.argv.slice(3));
             break;
         }
         case 'task-artifact': {
@@ -170,27 +170,32 @@ else
         }
         case 'task-review': {
             const { taskReview } = await import('../lib/internal/task-review.js');
-            taskReview(process.argv.slice(3));
+            await taskReview(process.argv.slice(3));
             break;
         }
         case 'task-event': {
             const { taskEvent } = await import('../lib/internal/task-event.js');
-            taskEvent(process.argv.slice(3));
+            await taskEvent(process.argv.slice(3));
+            break;
+        }
+        case 'task-invalidation': {
+            const { taskInvalidation } = await import('../lib/internal/task-invalidation.js');
+            taskInvalidation(process.argv.slice(3));
             break;
         }
         case 'task-lifecycle': {
             const { taskLifecycle } = await import('../lib/internal/task-lifecycle.js');
-            taskLifecycle(process.argv.slice(3));
+            await taskLifecycle(process.argv.slice(3));
             break;
         }
         case 'task-finalization': {
             const { taskFinalization } = await import('../lib/internal/task-finalization.js');
-            taskFinalization(process.argv.slice(3));
+            await taskFinalization(process.argv.slice(3));
             break;
         }
         case 'task-override': {
             const { taskOverride } = await import('../lib/internal/task-override.js');
-            taskOverride(process.argv.slice(3));
+            await taskOverride(process.argv.slice(3));
             break;
         }
         case 'task-short-id': {
@@ -205,7 +210,7 @@ else
         }
         case 'task-verify': {
             const { taskVerify } = await import('../lib/internal/task-verify.js');
-            taskVerify(process.argv.slice(3));
+            await taskVerify(process.argv.slice(3));
             break;
         }
         case 'task-validate': {

package/dist/lib/agent-clients/next-steps.js

@@ -1,6 +1,7 @@
 import { listEnabledAgentClientAdapters } from './registry.js';
 import { renderAgentClientInvocation } from './invocation.js';
 import semver from 'semver';
+import { isTaskScopeSkill } from '../task/skill-scope.js';
 const SAFE_NAME = /^[a-zA-Z0-9][a-zA-Z0-9._-]*$/;
 const SAFE_TASK_REF = /^(?:\d+|TASK-\d{8}-\d{6})$/;
 function requireSafeName(value, field) {
@@ -19,7 +20,11 @@ function renderNextStepCommands(input) {
         throw new Error('Invalid version');
     }
     const args = [
-        ...(input.taskRef === undefined ? [] : [input.taskRef]),
+        ...(input.taskRef === undefined
+            ? []
+            : isTaskScopeSkill(input.skillName)
+                ? ['--task', input.taskRef]
+                : [input.taskRef]),
         ...(input.version === undefined ? [] : [input.version])
     ];
     const builtins = listEnabledAgentClientAdapters(input.state).map((adapter) => Object.freeze({

package/dist/lib/decide.js

@@ -94,13 +94,13 @@ export async function decide(args, options = {}) {
             metadataProvider: () => ({ timestamp: now, agentInfraVersion: options.version ?? VERSION })
         };
         let result;
-        const execute = () => {
+        const execute = async () => {
             let current = applyHumanDecision(request, writeOptions);
             if (current.status !== 'failed' || !parsedDecision.overrideTicket)
                 return current;
             if (!parsedDecision.overrideTarget || !parsedDecision.overrideScope)
                 throw new Error('override ticket requires --override-target and --override-scope');
-            const consumed = consumeHumanOverride({
+            const consumed = await consumeHumanOverride({
                 taskRef: resolved.taskId,
                 ticketId: parsedDecision.overrideTicket,
                 failureId: failureId('decision-intent', current.error?.code ?? 'TASK_STATE_MISMATCH'),
@@ -120,7 +120,7 @@ export async function decide(args, options = {}) {
                 throw new Error(consumed.error.message);
             return current;
         };
-        result = withTaskExecutionLock(resolved.repoRoot, resolved.taskId, 'task-decision', execute);
+        result = await withTaskExecutionLock(resolved.repoRoot, resolved.taskId, 'task-decision', execute);
         if (result.error)
             throw new Error(result.error.message);
         return 0;

package/dist/lib/internal/platform-checks.js

@@ -76,7 +76,7 @@ async function platformChecks(args = []) {
         return;
     }
     if (operation === 'inspect') {
-        finish(inspectPullRequestReadiness(taskRef, { cwd }));
+        finish(await inspectPullRequestReadiness(taskRef, { cwd }));
         return;
     }
     if (operation === 'resolve-run') {
@@ -84,7 +84,7 @@ async function platformChecks(args = []) {
             fail('resolve-run requires --check-name');
             return;
         }
-        finish(resolvePlatformCheckRun(taskRef, { cwd, checkName: values.checkName, detailsUrl: values.detailsUrl }));
+        finish(await resolvePlatformCheckRun(taskRef, { cwd, checkName: values.checkName, detailsUrl: values.detailsUrl }));
         return;
     }
     if (operation === 'logs') {
@@ -94,7 +94,7 @@ async function platformChecks(args = []) {
             fail('logs requires a positive --run and optional positive --job');
             return;
         }
-        finish(fetchPlatformCheckLogs(taskRef, { cwd, run, job: job || undefined }));
+        finish(await fetchPlatformCheckLogs(taskRef, { cwd, run, job: job || undefined }));
         return;
     }
     const intervalSeconds = positive(values.intervalSeconds);

package/dist/lib/internal/platform-comment.js

@@ -3,7 +3,7 @@ import path from 'node:path';
 import { normalizeAgentToken, AGENT_USAGE_HINT } from '../agent-clients/tokens.js';
 import { checkPlatformCommentOwner, listPlatformComments, syncPlatformComment } from '../platform/issue-comments.js';
 import { backfillCompletionComments } from '../platform/completion-backfill.js';
-const USAGE = `Usage: agent-infra-internal platform-comment list --issue <N> [--cwd <path>]
+const USAGE = `Usage: agent-infra-internal platform-comment list --issue <token> [--cwd <path>]
        agent-infra-internal platform-comment owner <task-ref> [--cwd <path>]
        agent-infra-internal platform-comment backfill <task-ref> --agent <agent> [--cwd <path>]
        agent-infra-internal platform-comment sync <task-ref> --kind <kind> --agent <agent> [--artifact <file>] [--body-file <path|->] [--status-label <label>] [--backfill] [--cwd <path>]
@@ -47,7 +47,7 @@ function readBodyFile(value, cwd) {
         return fs.readFileSync(0, 'utf8');
     return fs.readFileSync(path.resolve(cwd, value), 'utf8');
 }
-function platformComment(args = []) {
+async function platformComment(args = []) {
     if (args[0] === '--help' || args[0] === '-h') {
         process.stdout.write(USAGE);
         return;
@@ -70,9 +70,9 @@ function platformComment(args = []) {
     }
     const cwd = path.resolve(typeof parsed.values.cwd === 'string' ? parsed.values.cwd : process.cwd());
     if (operation === 'list') {
-        const issue = Number(parsed.values.issue);
-        if (!Number.isInteger(issue) || issue <= 0) {
-            fail('list requires a positive --issue');
+        const issue = parsed.values.issue;
+        if (typeof issue !== 'string' || !issue) {
+            fail('list requires --issue <token>');
             return;
         }
         const unexpected = Object.keys(parsed.values).find((key) => !['issue', 'cwd'].includes(key));
@@ -80,7 +80,7 @@ function platformComment(args = []) {
             fail(`list does not accept '--${unexpected}'`);
             return;
         }
-        finish(listPlatformComments(issue, cwd));
+        finish(await listPlatformComments(issue, cwd));
         return;
     }
     if (operation === 'owner') {
@@ -89,7 +89,7 @@ function platformComment(args = []) {
             fail(`owner does not accept '--${unexpected}'`);
             return;
         }
-        finish(checkPlatformCommentOwner(taskRef, { cwd }));
+        finish(await checkPlatformCommentOwner(taskRef, { cwd }));
         return;
     }
     if (operation === 'backfill') {
@@ -108,7 +108,7 @@ function platformComment(args = []) {
             fail(`invalid --agent '${agent}': ${AGENT_USAGE_HINT}`);
             return;
         }
-        finish(backfillCompletionComments(taskRef, { cwd, agent: normalizedAgent }));
+        finish(await backfillCompletionComments(taskRef, { cwd, agent: normalizedAgent }));
         return;
     }
     const kind = parsed.values.kind;
@@ -143,7 +143,7 @@ function platformComment(args = []) {
         fail(`unable to read body file: ${error instanceof Error ? error.message : String(error)}`);
         return;
     }
-    finish(syncPlatformComment(taskRef, {
+    finish(await syncPlatformComment(taskRef, {
         kind: kind,
         agent: normalizedAgent,
         artifact: typeof parsed.values.artifact === 'string' ? parsed.values.artifact : undefined,

package/dist/lib/internal/platform-context.js

@@ -6,7 +6,7 @@ function fail(message) {
     process.stderr.write(USAGE);
     process.exitCode = 1;
 }
-function platformContext(args = []) {
+async function platformContext(args = []) {
     if (args[0] === '--help' || args[0] === '-h') {
         process.stdout.write(USAGE);
         return;
@@ -35,7 +35,7 @@ function platformContext(args = []) {
         seen.add(flag);
         cwd = path.resolve(value);
     }
-    const result = resolvePlatformContext({ cwd });
+    const result = await resolvePlatformContext({ cwd });
     process.stdout.write(`${JSON.stringify(result)}\n`);
     if (result.status === 'failed')
         process.exitCode = 1;

package/dist/lib/internal/platform-issue.js

@@ -3,7 +3,7 @@ import { normalizeAgentToken, AGENT_USAGE_HINT } from '../agent-clients/tokens.j
 import { bindPlatformIssue, createPlatformIssue, inspectPlatformIssue, syncPlatformIssue } from '../platform/issues.js';
 const USAGE = `Usage: agent-infra-internal platform-issue inspect <task-ref> [--cwd <path>]
        agent-infra-internal platform-issue create <task-ref> --agent <agent> [--dry-run] [--cwd <path>]
-       agent-infra-internal platform-issue bind <task-ref> --issue <N> --agent <agent> [--dry-run] [--cwd <path>]
+       agent-infra-internal platform-issue bind <task-ref> --issue <token> --agent <agent> [--dry-run] [--cwd <path>]
        agent-infra-internal platform-issue sync <task-ref> --agent <agent> [--status <suffix|none>] [--assignees current|none] [--milestone initial|specific|none] [--issue-type] [--fields] [--requirements] [--in-labels from-diff|none] [--base <branch>] [--state open|closed] [--close-reason completed|not_planned] [--dry-run] [--cwd <path>]
 `;
 const BOOLEAN_FLAGS = new Set(['--dry-run', '--issue-type', '--fields', '--requirements']);
@@ -54,7 +54,7 @@ function oneOf(values, name, allowed) {
         throw new Error(`--${name.replace(/[A-Z]/g, (letter) => `-${letter.toLowerCase()}`)} must be one of ${allowed.join('|')}`);
     return value;
 }
-function platformIssue(args = []) {
+async function platformIssue(args = []) {
     if (args[0] === '--help' || args[0] === '-h') {
         process.stdout.write(USAGE);
         return;
@@ -86,7 +86,7 @@ function platformIssue(args = []) {
         return;
     }
     if (operation === 'inspect') {
-        finish(inspectPlatformIssue(taskRef, { cwd }));
+        finish(await inspectPlatformIssue(taskRef, { cwd }));
         return;
     }
     if (typeof values.agent !== 'string' || !values.agent) {
@@ -100,16 +100,15 @@ function platformIssue(args = []) {
     }
     values.agent = agent;
     if (operation === 'create') {
-        finish(createPlatformIssue(taskRef, { cwd, agent: values.agent, dryRun: values.dryRun === true }));
+        finish(await createPlatformIssue(taskRef, { cwd, agent: values.agent, dryRun: values.dryRun === true }));
         return;
     }
     if (operation === 'bind') {
-        const issue = Number(values.issue);
-        if (!Number.isInteger(issue) || issue <= 0) {
-            fail('bind requires a positive --issue');
+        if (typeof values.issue !== 'string' || !values.issue) {
+            fail('bind requires --issue <token>');
             return;
         }
-        finish(bindPlatformIssue(taskRef, { cwd, agent: values.agent, issue, dryRun: values.dryRun === true }));
+        finish(await bindPlatformIssue(taskRef, { cwd, agent: values.agent, issue: values.issue, dryRun: values.dryRun === true }));
         return;
     }
     try {
@@ -127,7 +126,7 @@ function platformIssue(args = []) {
         const desiredKeys = ['status', 'assignees', 'milestone', 'issueType', 'fields', 'requirements', 'inLabels', 'state'];
         if (!desiredKeys.some((name) => values[name] !== undefined))
             throw new Error('sync requires at least one desired-state option');
-        finish(syncPlatformIssue(taskRef, {
+        finish(await syncPlatformIssue(taskRef, {
             cwd, agent: values.agent, dryRun: values.dryRun === true,
             status: typeof values.status === 'string' ? values.status : undefined,
             assignees: assignees,

package/dist/lib/internal/platform-pr-review.js

@@ -2,9 +2,9 @@ import fs from 'node:fs';
 import path from 'node:path';
 import { inspectPlatformPullRequestByNumber } from '../platform/pull-requests.js';
 import { listPrReviews, publishPrReview } from '../platform/pr-review.js';
-const USAGE = `Usage: agent-infra-internal platform-pr-review inspect --pr <N> [--cwd <path>]
-       agent-infra-internal platform-pr-review list --pr <N> [--cwd <path>]
-       agent-infra-internal platform-pr-review publish --pr <N> --scope <taskId|pr{N}> --round <N> --commit <sha> --event <COMMENT|APPROVE|REQUEST_CHANGES> --body-file <path|-> [--dry-run] [--cwd <path>]
+const USAGE = `Usage: agent-infra-internal platform-pr-review inspect --pr <token> [--cwd <path>]
+       agent-infra-internal platform-pr-review list --pr <token> [--cwd <path>]
+       agent-infra-internal platform-pr-review publish --pr <token> --scope <taskId|pr{N}> --round <N> --commit <sha> --event <COMMENT|APPROVE|REQUEST_CHANGES> --body-file <path|-> [--dry-run] [--cwd <path>]
 `;
 const BOOLEAN_FLAGS = new Set(['--dry-run']);
 const VALUE_FLAGS = new Set(['--cwd', '--pr', '--scope', '--round', '--commit', '--event', '--body-file']);
@@ -47,7 +47,7 @@ function parse(args, start) {
 function readBodyFile(value, cwd) {
     return value === '-' ? fs.readFileSync(0, 'utf8') : fs.readFileSync(path.resolve(cwd, value), 'utf8');
 }
-function platformPrReview(args = []) {
+async function platformPrReview(args = []) {
     if (args[0] === '--help' || args[0] === '-h') {
         process.stdout.write(USAGE);
         return;
@@ -74,17 +74,17 @@ function platformPrReview(args = []) {
         fail(`${operation} does not accept --${unexpected}`);
         return;
     }
-    const pr = Number(values.pr);
-    if (!Number.isInteger(pr) || pr <= 0) {
-        fail(`${operation} requires a positive --pr`);
+    const pr = values.pr;
+    if (typeof pr !== 'string' || !pr) {
+        fail(`${operation} requires --pr <token>`);
         return;
     }
     if (operation === 'inspect') {
-        finish(inspectPlatformPullRequestByNumber(pr, { cwd }));
+        finish(await inspectPlatformPullRequestByNumber(pr, { cwd }));
         return;
     }
     if (operation === 'list') {
-        finish(listPrReviews(pr, { cwd }));
+        finish(await listPrReviews(pr, { cwd }));
         return;
     }
     const scope = typeof values.scope === 'string' ? values.scope : '';
@@ -119,7 +119,7 @@ function platformPrReview(args = []) {
         fail(`unable to read body file: ${error instanceof Error ? error.message : String(error)}`);
         return;
     }
-    finish(publishPrReview({
+    finish(await publishPrReview({
         cwd,
         prNumber: pr,
         identity: { scope, round, commitSha: commit },