@swarmclawai/swarmclaw 1.9.31 → 1.9.33

package/README.md

@@ -146,10 +146,29 @@ Then open `http://localhost:3456`.
 Install the SwarmClaw skill for your [OpenClaw](https://openclaw.ai) agents:
 
 ```bash
-clawhub install swarmclaw
+openclaw skills install swarmclaw
 ```
 
-[Browse on ClawHub](https://clawhub.ai/skills/swarmclaw)
+[Browse on ClawHub](https://clawhub.ai/waydelyle/swarmclaw)
+
+## v1.9.33 Highlights
+
+Issue and PR validation release for credential durability, delegated task dispatch, connector output hygiene, and OpenClaw gateway protocol compatibility.
+
+- **Credential durability.** Execute-tool credential injection now reads the persisted `encryptedKey` field, and `CREDENTIAL_SECRET` now resolves in a stable order: explicit environment value, `DATA_DIR/credential-secret`, legacy env files, then generated fallback.
+- **Delegated task dispatch.** Agent-created tasks delegated to another agent auto-queue when no explicit status is supplied, and failed dead-lettered tasks can be requeued through `POST /api/tasks/:id/retry`.
+- **Connector output hygiene.** Connector replies now reuse the internal metadata scrubber before delivery and persistence, while successful non-connector delivery tool output is no longer overwritten as an unconfirmed send.
+- **Agent and gateway compatibility.** Agent updates preserve workspace filesystem settings, and OpenClaw gateway routes now use protocol version 4.
+- **Regression coverage.** Added tests for credential env injection, secret precedence, delegated queueing, failed-task retry, connector sanitization, agent workspace settings, and OpenClaw gateway protocol exports.
+
+## v1.9.32 Highlights
+
+PR integration release for background model routing, reflection memory controls, and current ClawHub install guidance.
+
+- **Background model routing.** Per-agent `dreamConfig` overrides can route dream cycles and daily digests before global dream settings, while `compactionProvider` settings can route live auto-compaction summaries through a cheaper or faster model.
+- **Reflection memory controls.** `reflectionMinQuality` gates automatic reflection memory writes without dropping the reflection record, and optional embedding dedup skips near-duplicate reflection notes when embeddings are configured.
+- **ClawHub install guidance.** OpenClaw skill docs now use `openclaw skills install swarmclaw` and current owner-scoped ClawHub links.
+- **Regression coverage.** Added tests for dream override precedence, compaction preference resolution, reflection quality gating, and embedding-based reflection dedup.
 
 ## v1.9.31 Highlights
 
@@ -403,7 +422,7 @@ SwarmClaw agents can join [SwarmFeed](https://swarmfeed.ai) — a social network
 - **Per-agent opt-in**: enable SwarmFeed on any agent with automatic Ed25519 registration
 - **Richer in-app surface**: feed tabs for For You, Following, Trending, Bookmarks, and Notifications, plus thread detail, profile sheets, suggested follows, and search
 - **Heartbeat integration**: agents can auto-post, auto-reply to mentions, auto-follow with guardrails, and publish task-completion updates during heartbeat cycles
-- **Multiple access methods**: [SDK](https://www.npmjs.com/package/@swarmfeed/sdk), [CLI](https://www.npmjs.com/package/@swarmfeed/cli), [MCP Server](https://www.npmjs.com/package/@swarmfeed/mcp-server), and [ClawHub skill](https://clawhub.ai/skills/swarmfeed)
+- **Multiple access methods**: [SDK](https://www.npmjs.com/package/@swarmfeed/sdk), [CLI](https://www.npmjs.com/package/@swarmfeed/cli), [MCP Server](https://www.npmjs.com/package/@swarmfeed/mcp-server), and [ClawHub skill](https://clawhub.ai/waydelyle/swarmfeed)
 
 Read the docs at [swarmclaw.ai/docs/swarmfeed](https://swarmclaw.ai/docs/swarmfeed) and visit [swarmfeed.ai](https://swarmfeed.ai) for the platform itself.
 
@@ -426,6 +445,25 @@ Operational docs: https://swarmclaw.ai/docs/observability
 
 ## Releases
 
+### v1.9.33 Highlights
+
+Issue and PR validation release for credential durability, delegated task dispatch, connector output hygiene, and OpenClaw gateway protocol compatibility.
+
+- **Credential durability.** Execute-tool credential injection now reads the persisted `encryptedKey` field, and `CREDENTIAL_SECRET` now resolves in a stable order: explicit environment value, `DATA_DIR/credential-secret`, legacy env files, then generated fallback.
+- **Delegated task dispatch.** Agent-created tasks delegated to another agent auto-queue when no explicit status is supplied, and failed dead-lettered tasks can be requeued through `POST /api/tasks/:id/retry`.
+- **Connector output hygiene.** Connector replies now reuse the internal metadata scrubber before delivery and persistence, while successful non-connector delivery tool output is no longer overwritten as an unconfirmed send.
+- **Agent and gateway compatibility.** Agent updates preserve workspace filesystem settings, and OpenClaw gateway routes now use protocol version 4.
+- **Regression coverage.** Added tests for credential env injection, secret precedence, delegated queueing, failed-task retry, connector sanitization, agent workspace settings, and OpenClaw gateway protocol exports.
+
+### v1.9.32 Highlights
+
+PR integration release for background model routing, reflection memory controls, and current ClawHub install guidance.
+
+- **Background model routing.** Per-agent `dreamConfig` overrides can route dream cycles and daily digests before global dream settings, while `compactionProvider` settings can route live auto-compaction summaries through a cheaper or faster model.
+- **Reflection memory controls.** `reflectionMinQuality` gates automatic reflection memory writes without dropping the reflection record, and optional embedding dedup skips near-duplicate reflection notes when embeddings are configured.
+- **ClawHub install guidance.** OpenClaw skill docs now use `openclaw skills install swarmclaw` and current owner-scoped ClawHub links.
+- **Regression coverage.** Added tests for dream override precedence, compaction preference resolution, reflection quality gating, and embedding-based reflection dedup.
+
 ### v1.9.31 Highlights
 
 Documentation cleanup release for public release notes and OpenClaw guidance. No runtime behavior changed.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@swarmclawai/swarmclaw",
-  "version": "1.9.31",
+  "version": "1.9.33",
   "description": "Build and run autonomous AI agents with OpenClaw, Hermes, multiple model providers, orchestration, delegation, memory, skills, schedules, and chat connectors.",
   "main": "electron-dist/main.js",
   "license": "MIT",
@@ -88,7 +88,7 @@
     "test:cli": "node --test src/cli/*.test.js bin/*.test.js scripts/electron-after-pack.test.mjs scripts/electron-signing-config.test.mjs scripts/ensure-sandbox-browser-image.test.mjs scripts/postinstall.test.mjs scripts/run-next-build.test.mjs scripts/run-next-typegen.test.mjs",
     "test:setup": "tsx --test src/app/api/setup/check-provider/route.test.ts src/lib/server/provider-model-discovery.test.ts src/components/auth/setup-wizard/utils.test.ts src/components/auth/setup-wizard/types.test.ts src/hooks/setup-done-detection.test.ts src/lib/setup-defaults.test.ts src/lib/server/storage-auth.test.ts src/lib/server/storage-auth-docker.test.ts",
     "test:openclaw": "tsx --test src/lib/openclaw/openclaw-agent-id.test.ts src/lib/openclaw/openclaw-endpoint.test.ts src/lib/server/agents/agent-runtime-config.test.ts src/lib/server/build-llm.test.ts src/lib/server/connectors/connector-routing.test.ts src/lib/server/connectors/openclaw.test.ts src/lib/server/connectors/swarmdock.test.ts src/lib/server/gateway/protocol.test.ts src/lib/server/gateways/gateway-topology.test.ts src/lib/server/llm-response-cache.test.ts src/lib/server/mcp-conformance.test.ts src/lib/server/openclaw/agent-resolver.test.ts src/lib/server/openclaw/deploy.test.ts src/lib/server/openclaw/skills-normalize.test.ts src/lib/server/session-tools/openclaw-nodes.test.ts src/lib/server/session-tools/swarmdock.test.ts src/lib/server/tasks/task-quality-gate.test.ts src/lib/server/tasks/task-validation.test.ts src/lib/server/tool-capability-policy.test.ts src/lib/providers/openai.test.ts src/lib/providers/openclaw-exports.test.ts src/app/api/gateways/topology-route.test.ts src/app/api/openclaw/dashboard-url/route.test.ts",
-    "test:runtime": "tsx --test src/lib/a2a/agent-card.test.ts src/lib/agent-planning-mode.test.ts src/lib/agent-config-history.test.ts src/lib/strip-internal-metadata.test.ts src/lib/provider-sets.test.ts src/lib/providers/opencode-cli.test.ts src/lib/providers/cli-provider-metadata.test.ts src/lib/providers/cli-utils.test.ts src/lib/providers/generic-cli.test.ts src/lib/server/agents/delegation-advisory.test.ts src/lib/server/agents/agent-runtime-config.test.ts src/lib/server/cli-provider-readiness.test.ts src/lib/server/provider-health.test.ts src/lib/server/provider-diagnostics.test.ts src/lib/server/mcp-gateway-runtime.test.ts src/lib/server/mcp-connection-pool.test.ts src/lib/server/knowledge-sources.test.ts src/lib/server/memory/dream-service.test.ts src/lib/server/extension-managed-resources.test.ts src/lib/server/eval/baseline.test.ts src/lib/server/eval/environment-plan.test.ts src/lib/server/chat-execution/chat-execution-grounding.test.ts src/lib/server/chat-execution/chat-turn-preparation.test.ts src/lib/server/chat-execution/iteration-timers.test.ts src/lib/server/chat-execution/post-stream-finalization.test.ts src/lib/server/chat-execution/prompt-sections.planning-mode.test.ts src/lib/server/chat-execution/reasoning-tag-scrubber.test.ts src/lib/server/chats/clear-undo-snapshots.test.ts src/lib/server/chats/session-context-pack.test.ts src/lib/server/connectors/email.test.ts src/lib/server/connectors/slack.test.ts src/lib/server/protocols/protocol-service.test.ts src/lib/server/runtime/run-ledger.test.ts src/lib/server/runtime/queue-retry-policy.test.ts src/lib/server/runs/run-brief.test.ts src/lib/server/runs/run-handoff.test.ts src/lib/server/operations/operation-pulse.test.ts src/lib/server/schedules/schedule-history.test.ts src/lib/server/schedules/schedule-timing.test.ts src/lib/server/schedules/schedule-preview.test.ts src/lib/quality/release-readiness.test.ts src/lib/quality/architecture-health.test.ts src/lib/server/artifacts/artifact-resolver.test.ts src/lib/server/observability/otel-config.test.ts src/lib/server/safe-parse-body.test.ts src/lib/server/missions/mission-templates.test.ts src/lib/server/sharing/share-link-repository.test.ts src/lib/server/sharing/share-resolver.test.ts src/lib/server/tasks/task-execution-workspace.test.ts src/lib/server/tasks/task-execution-policy.test.ts src/lib/server/tasks/task-handoff.test.ts src/lib/server/tasks/task-service.test.ts src/lib/server/session-tools/execute.test.ts src/lib/server/session-tools/manage-tasks.test.ts src/lib/server/session-tools/web-crawl.test.ts src/lib/app/view-constants.test.ts src/lib/quality/quality-summary.test.ts src/app/api/approvals/route.test.ts src/app/api/agents/agents-route.test.ts src/app/api/tasks/tasks-route.test.ts src/app/api/tasks/task-workspace-route.test.ts src/app/api/chats/chat-route.test.ts src/app/api/chats/clear-route.test.ts src/app/api/chats/compact-route.test.ts src/app/api/chats/context-pack-route.test.ts src/app/api/chats/context-status-route.test.ts src/app/api/config-versions/config-versions-route.test.ts src/app/api/runs/run-handoff-route.test.ts src/app/api/connectors/connector-doctor-route.test.ts src/app/api/extensions/managed-resources/route.test.ts src/app/api/gateways/control-route.test.ts src/app/api/healthz/route.test.ts src/app/api/logs/route.test.ts src/app/api/portability/export/route.test.ts src/app/api/portability/import/route.test.ts src/app/api/providers/[id]/route.test.ts src/app/api/schedules/preview/route.test.ts src/app/api/schedules/schedule-history-route.test.ts src/app/api/tts/route.test.ts",
+    "test:runtime": "tsx --test src/lib/a2a/agent-card.test.ts src/lib/agent-planning-mode.test.ts src/lib/agent-config-history.test.ts src/lib/autonomy/supervisor-settings.test.ts src/lib/strip-internal-metadata.test.ts src/lib/provider-sets.test.ts src/lib/providers/opencode-cli.test.ts src/lib/providers/cli-provider-metadata.test.ts src/lib/providers/cli-utils.test.ts src/lib/providers/generic-cli.test.ts src/lib/server/agents/delegation-advisory.test.ts src/lib/server/agents/agent-runtime-config.test.ts src/lib/server/autonomy/supervisor-reflection.test.ts src/lib/server/cli-provider-readiness.test.ts src/lib/server/provider-health.test.ts src/lib/server/provider-diagnostics.test.ts src/lib/server/mcp-gateway-runtime.test.ts src/lib/server/mcp-connection-pool.test.ts src/lib/server/knowledge-sources.test.ts src/lib/server/memory/dream-service.test.ts src/lib/server/memory/memory-consolidation.test.ts src/lib/server/extension-managed-resources.test.ts src/lib/server/eval/baseline.test.ts src/lib/server/eval/environment-plan.test.ts src/lib/server/chat-execution/chat-execution-grounding.test.ts src/lib/server/chat-execution/chat-turn-preparation.test.ts src/lib/server/chat-execution/compaction-generation-preference.test.ts src/lib/server/chat-execution/iteration-timers.test.ts src/lib/server/chat-execution/post-stream-finalization.test.ts src/lib/server/chat-execution/prompt-sections.planning-mode.test.ts src/lib/server/chat-execution/reasoning-tag-scrubber.test.ts src/lib/server/chats/clear-undo-snapshots.test.ts src/lib/server/chats/session-context-pack.test.ts src/lib/server/connectors/email.test.ts src/lib/server/connectors/slack.test.ts src/lib/server/protocols/protocol-service.test.ts src/lib/server/runtime/run-ledger.test.ts src/lib/server/runtime/queue-retry-policy.test.ts src/lib/server/runs/run-brief.test.ts src/lib/server/runs/run-handoff.test.ts src/lib/server/operations/operation-pulse.test.ts src/lib/server/schedules/schedule-history.test.ts src/lib/server/schedules/schedule-timing.test.ts src/lib/server/schedules/schedule-preview.test.ts src/lib/quality/release-readiness.test.ts src/lib/quality/architecture-health.test.ts src/lib/server/artifacts/artifact-resolver.test.ts src/lib/server/observability/otel-config.test.ts src/lib/server/safe-parse-body.test.ts src/lib/server/missions/mission-templates.test.ts src/lib/server/sharing/share-link-repository.test.ts src/lib/server/sharing/share-resolver.test.ts src/lib/server/tasks/task-execution-workspace.test.ts src/lib/server/tasks/task-execution-policy.test.ts src/lib/server/tasks/task-handoff.test.ts src/lib/server/tasks/task-service.test.ts src/lib/server/session-tools/execute.test.ts src/lib/server/session-tools/manage-tasks.test.ts src/lib/server/session-tools/web-crawl.test.ts src/lib/app/view-constants.test.ts src/lib/quality/quality-summary.test.ts src/app/api/approvals/route.test.ts src/app/api/agents/agents-route.test.ts src/app/api/tasks/tasks-route.test.ts src/app/api/tasks/task-workspace-route.test.ts src/app/api/chats/chat-route.test.ts src/app/api/chats/clear-route.test.ts src/app/api/chats/compact-route.test.ts src/app/api/chats/context-pack-route.test.ts src/app/api/chats/context-status-route.test.ts src/app/api/config-versions/config-versions-route.test.ts src/app/api/runs/run-handoff-route.test.ts src/app/api/connectors/connector-doctor-route.test.ts src/app/api/extensions/managed-resources/route.test.ts src/app/api/gateways/control-route.test.ts src/app/api/healthz/route.test.ts src/app/api/logs/route.test.ts src/app/api/portability/export/route.test.ts src/app/api/portability/import/route.test.ts src/app/api/providers/[id]/route.test.ts src/app/api/schedules/preview/route.test.ts src/app/api/schedules/schedule-history-route.test.ts src/app/api/tts/route.test.ts",
     "test:builder": "tsx --test src/features/protocols/builder/utils/nodes-to-template.test.ts src/features/protocols/builder/utils/template-to-nodes.test.ts src/features/protocols/builder/validators/dag-validator.test.ts",
     "test:e2e": "node --import tsx scripts/browser-e2e-smoke.ts",
     "test:mcp:conformance": "node --import tsx ./scripts/mcp-conformance-check.ts",
@@ -154,7 +154,7 @@
     "next": "16.2.4",
     "next-themes": "^0.4.6",
     "nodemailer": "^8.0.1",
-    "openclaw": "^2026.4.22",
+    "openclaw": "^2026.5.12",
     "pdf-parse": "^2.4.5",
     "qrcode": "^1.5.4",
     "radix-ui": "^1.4.3",

package/src/app/api/agents/agents-route.test.ts

@@ -206,6 +206,42 @@ test('PUT /api/agents/:id updates planning mode without clobbering other fields'
   assert.equal(body.proactiveMemory, false)
 })
 
+test('PUT /api/agents/:id persists workspace filesystem settings', async () => {
+  seedAgent('agent-workspace-settings', {
+    name: 'Workspace Agent',
+    workspace: null,
+    filesystemScope: null,
+    fileAccessPolicy: null,
+  })
+
+  const response = await putAgent(new Request('http://local/api/agents/agent-workspace-settings', {
+    method: 'PUT',
+    headers: { 'content-type': 'application/json' },
+    body: JSON.stringify({
+      workspace: '/tmp/swarmclaw-agent-workspace',
+      filesystemScope: 'workspace',
+      fileAccessPolicy: {
+        allowedPaths: ['/tmp/swarmclaw-agent-workspace'],
+        blockedPaths: ['/tmp/swarmclaw-agent-workspace/private'],
+      },
+    }),
+  }), routeParams('agent-workspace-settings'))
+
+  assert.equal(response.status, 200)
+  const body = await response.json()
+  assert.equal(body.workspace, '/tmp/swarmclaw-agent-workspace')
+  assert.equal(body.filesystemScope, 'workspace')
+  assert.deepEqual(body.fileAccessPolicy, {
+    allowedPaths: ['/tmp/swarmclaw-agent-workspace'],
+    blockedPaths: ['/tmp/swarmclaw-agent-workspace/private'],
+  })
+
+  const stored = loadAgents()['agent-workspace-settings']
+  assert.equal(stored.workspace, '/tmp/swarmclaw-agent-workspace')
+  assert.equal(stored.filesystemScope, 'workspace')
+  assert.deepEqual(stored.fileAccessPolicy, body.fileAccessPolicy)
+})
+
 test('PUT /api/agents/:id rejects non-string name', async () => {
   seedAgent('agent-bad-name', { name: 'Good' })
 

package/src/app/api/tasks/[id]/retry/route.ts

@@ -0,0 +1,12 @@
+import { NextResponse } from 'next/server'
+import { notFound } from '@/lib/server/collection-helpers'
+import { retryTaskFromRoute } from '@/lib/server/tasks/task-route-service'
+
+export async function POST(_req: Request, { params }: { params: Promise<{ id: string }> }) {
+  const { id } = await params
+  const result = retryTaskFromRoute(id)
+  if (!result.ok && result.status === 404) return notFound()
+  return result.ok
+    ? NextResponse.json(result.payload)
+    : NextResponse.json(result.payload, { status: result.status })
+}

package/src/app/api/tasks/task-workspace-route.test.ts

@@ -19,6 +19,7 @@ let getTaskHandoff: typeof import('./[id]/handoff/route')['GET']
 let postTaskHandoff: typeof import('./[id]/handoff/route')['POST']
 let getTaskExecutionPolicy: typeof import('./[id]/execution-policy/route')['GET']
 let postTaskExecutionPolicy: typeof import('./[id]/execution-policy/route')['POST']
+let postTaskRetry: typeof import('./[id]/retry/route')['POST']
 let getTaskHandoffs: typeof import('./handoffs/route')['GET']
 let getTasks: typeof import('./route')['GET']
 let storage: typeof import('@/lib/server/storage')
@@ -57,6 +58,7 @@ before(async () => {
   const policyRoute = await import('./[id]/execution-policy/route')
   getTaskExecutionPolicy = policyRoute.GET
   postTaskExecutionPolicy = policyRoute.POST
+  postTaskRetry = (await import('./[id]/retry/route')).POST
   getTaskHandoffs = (await import('./handoffs/route')).GET
   getTasks = (await import('./route')).GET
 })
@@ -255,6 +257,66 @@ test('GET /api/tasks/:id/execution-policy returns policy summary', async () => {
   assert.equal(body.summary.status, 'waiting')
 })
 
+test('POST /api/tasks/:id/retry requeues a dead-lettered failed task', async () => {
+  seedTask('task-dead-letter-retry', {
+    title: 'Dead Letter Retry',
+    status: 'failed',
+    attempts: 3,
+    maxAttempts: 3,
+    retryScheduledAt: Date.now() + 60_000,
+    deadLetteredAt: Date.now(),
+    checkoutRunId: 'run-failed',
+    error: 'Dead-lettered after 3/3 attempts: timeout',
+    validation: { ok: false, reasons: ['No result'], checkedAt: Date.now() },
+  })
+
+  const response = await postTaskRetry(
+    new Request('http://local/api/tasks/task-dead-letter-retry/retry', { method: 'POST' }),
+    routeParams('task-dead-letter-retry'),
+  )
+
+  assert.equal(response.status, 200)
+  const body = await response.json() as BoardTask
+  assert.equal(body.status, 'queued')
+  assert.equal(body.attempts, 0)
+  assert.equal(body.retryScheduledAt, null)
+  assert.equal(body.deadLetteredAt, null)
+  assert.equal(body.checkoutRunId, null)
+  assert.equal(body.error, null)
+  assert.equal(body.validation, null)
+  assert.equal(storage.loadQueue().includes('task-dead-letter-retry'), true)
+  assert.equal(body.comments?.some((comment) => comment.text.includes('retry requested')), true)
+})
+
+test('POST /api/tasks/:id/retry rejects tasks still blocked by dependencies', async () => {
+  seedTask('task-blocked-retry', {
+    title: 'Blocked Retry',
+    status: 'failed',
+    blockedBy: ['retry-dep'],
+    deadLetteredAt: Date.now(),
+  })
+  const tasks = storage.loadTasks()
+  tasks['retry-dep'] = {
+    id: 'retry-dep',
+    title: 'Retry Dependency',
+    description: '',
+    status: 'running',
+    agentId: 'agent-1',
+    createdAt: Date.now(),
+    updatedAt: Date.now(),
+  } as BoardTask
+  storage.saveTasks(tasks)
+
+  const response = await postTaskRetry(
+    new Request('http://local/api/tasks/task-blocked-retry/retry', { method: 'POST' }),
+    routeParams('task-blocked-retry'),
+  )
+
+  assert.equal(response.status, 409)
+  assert.equal(storage.loadTasks()['task-blocked-retry']?.status, 'failed')
+  assert.equal(storage.loadQueue().includes('task-blocked-retry'), false)
+})
+
 test('POST /api/tasks/:id/handoff saves markdown and JSON snapshots into the workspace', async () => {
   seedTask('task-handoff-save', {
     title: 'Saved Handoff Task',

package/src/cli/index.js

@@ -761,6 +761,7 @@ const COMMAND_GROUPS = [
       cmd('handoffs', 'GET', '/tasks/handoffs', 'List task handoff readiness packets'),
       cmd('execution-policy', 'GET', '/tasks/:id/execution-policy', 'Get task execution policy state'),
       cmd('execution-policy-decision', 'POST', '/tasks/:id/execution-policy', 'Approve, request changes, or reset a task policy stage', { expectsJsonBody: true }),
+      cmd('retry', 'POST', '/tasks/:id/retry', 'Retry a failed task and requeue it'),
       cmd('create', 'POST', '/tasks', 'Create task', { expectsJsonBody: true }),
       cmd('bulk', 'POST', '/tasks/bulk', 'Bulk update tasks (status/agent/project)', { expectsJsonBody: true }),
       cmd('update', 'PUT', '/tasks/:id', 'Update task', { expectsJsonBody: true }),

package/src/cli/index.test.js

@@ -225,6 +225,36 @@ test('tasks execution-policy-decision posts policy decisions', async () => {
   assert.equal(stderr.toString(), '')
 })
 
+test('tasks retry posts to the failed-task retry endpoint', async () => {
+  const stdout = makeWritable()
+  const stderr = makeWritable()
+  const calls = []
+
+  const fetchImpl = async (url, init) => {
+    calls.push({ url: String(url), init })
+    return jsonResponse({ id: 'task-1', status: 'queued' })
+  }
+
+  const exitCode = await runCli(
+    ['tasks', 'retry', 'task-1', '--json'],
+    {
+      fetchImpl,
+      stdout,
+      stderr,
+      env: {},
+      cwd: process.cwd(),
+    }
+  )
+
+  assert.equal(exitCode, 0)
+  assert.equal(calls.length, 1)
+  assert.match(calls[0].url, /\/api\/tasks\/task-1\/retry$/)
+  assert.equal(calls[0].init.method, 'POST')
+  assert.equal(calls[0].init.body, undefined)
+  assert.match(stdout.toString(), /"queued"/)
+  assert.equal(stderr.toString(), '')
+})
+
 test('gateways drain command posts a lifecycle control action', async () => {
   const stdout = makeWritable()
   const stderr = makeWritable()

package/src/cli/spec.js

@@ -541,6 +541,7 @@ const COMMAND_GROUPS = {
       handoffs: { description: 'List task handoff readiness packets', method: 'GET', path: '/tasks/handoffs' },
       'execution-policy': { description: 'Get task execution policy state', method: 'GET', path: '/tasks/:id/execution-policy', params: ['id'] },
       'execution-policy-decision': { description: 'Approve, request changes, or reset a task policy stage', method: 'POST', path: '/tasks/:id/execution-policy', params: ['id'] },
+      retry: { description: 'Retry a failed task and requeue it', method: 'POST', path: '/tasks/:id/retry', params: ['id'] },
       create: { description: 'Create task', method: 'POST', path: '/tasks' },
       bulk: { description: 'Bulk update tasks (status/agent/project)', method: 'POST', path: '/tasks/bulk' },
       update: { description: 'Update task', method: 'PUT', path: '/tasks/:id', params: ['id'] },

package/src/lib/autonomy/supervisor-settings.test.ts

@@ -0,0 +1,26 @@
+import assert from 'node:assert/strict'
+import { describe, it } from 'node:test'
+import { normalizeSupervisorSettings } from './supervisor-settings'
+
+describe('normalizeSupervisorSettings', () => {
+  it('preserves and clamps reflection memory quality settings', () => {
+    assert.deepEqual(
+      {
+        minQuality: normalizeSupervisorSettings({ reflectionMinQuality: '0.72' }).reflectionMinQuality,
+        minQualityHigh: normalizeSupervisorSettings({ reflectionMinQuality: 2 }).reflectionMinQuality,
+        minQualityLow: normalizeSupervisorSettings({ reflectionMinQuality: -1 }).reflectionMinQuality,
+        semanticEnabled: normalizeSupervisorSettings({ reflectionSemanticDedupEnabled: 'on' }).reflectionSemanticDedupEnabled,
+        semanticThreshold: normalizeSupervisorSettings({ reflectionSemanticDedupThreshold: '0.91' }).reflectionSemanticDedupThreshold,
+        semanticThresholdHigh: normalizeSupervisorSettings({ reflectionSemanticDedupThreshold: 1.5 }).reflectionSemanticDedupThreshold,
+      },
+      {
+        minQuality: 0.72,
+        minQualityHigh: 1,
+        minQualityLow: 0,
+        semanticEnabled: true,
+        semanticThreshold: 0.91,
+        semanticThresholdHigh: 1,
+      },
+    )
+  })
+})

package/src/lib/autonomy/supervisor-settings.ts

@@ -8,6 +8,9 @@ export const DEFAULT_SUPERVISOR_NO_PROGRESS_LIMIT = 2
 export const DEFAULT_SUPERVISOR_REPEATED_TOOL_LIMIT = 3
 export const DEFAULT_REFLECTION_ENABLED = true
 export const DEFAULT_REFLECTION_AUTO_WRITE_MEMORY = true
+export const DEFAULT_REFLECTION_MIN_QUALITY = 0
+export const DEFAULT_REFLECTION_SEMANTIC_DEDUP_ENABLED = false
+export const DEFAULT_REFLECTION_SEMANTIC_DEDUP_THRESHOLD = 0.88
 
 export const SUPERVISOR_NO_PROGRESS_LIMIT_MIN = 1
 export const SUPERVISOR_NO_PROGRESS_LIMIT_MAX = 8
@@ -24,6 +27,16 @@ function parseIntSetting(value: unknown, fallback: number, min: number, max: num
   return Math.max(min, Math.min(max, Math.trunc(parsed)))
 }
 
+function parseNumberSetting(value: unknown, fallback: number, min: number, max: number): number {
+  const parsed = typeof value === 'number'
+    ? value
+    : typeof value === 'string'
+      ? Number.parseFloat(value)
+      : Number.NaN
+  if (!Number.isFinite(parsed)) return fallback
+  return Math.max(min, Math.min(max, parsed))
+}
+
 function parseBoolSetting(value: unknown, fallback: boolean): boolean {
   if (typeof value === 'boolean') return value
   if (typeof value === 'string') {
@@ -41,6 +54,9 @@ export interface NormalizedSupervisorSettings {
   supervisorRepeatedToolLimit: number
   reflectionEnabled: boolean
   reflectionAutoWriteMemory: boolean
+  reflectionMinQuality: number
+  reflectionSemanticDedupEnabled: boolean
+  reflectionSemanticDedupThreshold: number
 }
 
 export function normalizeSupervisorSettings(
@@ -69,6 +85,17 @@ export function normalizeSupervisorSettings(
     ),
     reflectionEnabled: parseBoolSetting(current.reflectionEnabled, DEFAULT_REFLECTION_ENABLED),
     reflectionAutoWriteMemory: parseBoolSetting(current.reflectionAutoWriteMemory, DEFAULT_REFLECTION_AUTO_WRITE_MEMORY),
+    reflectionMinQuality: parseNumberSetting(current.reflectionMinQuality, DEFAULT_REFLECTION_MIN_QUALITY, 0, 1),
+    reflectionSemanticDedupEnabled: parseBoolSetting(
+      current.reflectionSemanticDedupEnabled,
+      DEFAULT_REFLECTION_SEMANTIC_DEDUP_ENABLED,
+    ),
+    reflectionSemanticDedupThreshold: parseNumberSetting(
+      current.reflectionSemanticDedupThreshold,
+      DEFAULT_REFLECTION_SEMANTIC_DEDUP_THRESHOLD,
+      0,
+      1,
+    ),
   }
 }
 

package/src/lib/providers/openclaw.test.ts

@@ -1,7 +1,7 @@
 import assert from 'node:assert/strict'
 import { afterEach, test } from 'node:test'
 
-import { buildOpenClawSessionKey, resolveGatewayAgentId } from './openclaw'
+import { buildOpenClawConnectParams, buildOpenClawSessionKey, resolveGatewayAgentId } from './openclaw'
 import { loadAgents, saveAgents } from '../server/storage'
 import type { Agent } from '@/types'
 
@@ -72,3 +72,10 @@ test('buildOpenClawSessionKey honors explicit OpenClaw session keys when provide
 
   assert.equal(sessionKey, 'agent:ops:benchmark:fixed-key')
 })
+
+test('buildOpenClawConnectParams advertises the current gateway protocol', () => {
+  const params = buildOpenClawConnectParams('test-token', 'test-nonce')
+
+  assert.equal(params.minProtocol, 4)
+  assert.equal(params.maxProtocol, 4)
+})

package/src/lib/providers/openclaw.ts

@@ -113,6 +113,8 @@ export function getDeviceId(): string {
 
 // --- Protocol helpers ---
 
+export const OPENCLAW_GATEWAY_PROTOCOL_VERSION = 4
+
 /**
  * Build connect params for the OpenClaw gateway protocol.
  *
@@ -132,8 +134,8 @@ export function buildOpenClawConnectParams(
   const scopes = ['operator.admin']
 
   const params: Record<string, unknown> = {
-    minProtocol: 3,
-    maxProtocol: 3,
+    minProtocol: OPENCLAW_GATEWAY_PROTOCOL_VERSION,
+    maxProtocol: OPENCLAW_GATEWAY_PROTOCOL_VERSION,
     auth: token ? { token } : undefined,
     client: {
       id: clientId,

package/src/lib/server/autonomy/supervisor-reflection.test.ts

@@ -202,6 +202,212 @@ describe('supervisor-reflection', () => {
     ])
   })
 
+  it('persists low-quality reflections while skipping auto-written memory', () => {
+    const output = runWithTempDataDir(`
+      const storageMod = await import('@/lib/server/storage')
+      const storage = storageMod.default || storageMod['module.exports'] || storageMod
+      const reflectionMod = await import('@/lib/server/autonomy/supervisor-reflection')
+      const mod = reflectionMod.default || reflectionMod['module.exports'] || reflectionMod
+      const memoryDbMod = await import('@/lib/server/memory/memory-db')
+      const memoryMod = memoryDbMod.default || memoryDbMod['module.exports'] || memoryDbMod
+
+      storage.saveAgents({
+        'agent-a': {
+          id: 'agent-a',
+          name: 'Agent A',
+          provider: 'openai',
+          model: 'gpt-test',
+        },
+      })
+
+      storage.saveSessions({
+        s1: {
+          id: 's1',
+          name: 'Autonomy Session',
+          cwd: process.cwd(),
+          user: 'tester',
+          provider: 'openai',
+          model: 'gpt-test',
+          claudeSessionId: null,
+          messages: [
+            { role: 'user', text: 'Repair the deployment workflow and keep notes for later.', time: 1 },
+            { role: 'assistant', text: 'I retried the same shell path and nothing changed.', time: 2 },
+          ],
+          createdAt: 1,
+          lastActiveAt: 2,
+          sessionType: 'human',
+          agentId: 'agent-a',
+        },
+      })
+
+      storage.saveSettings({
+        supervisorEnabled: true,
+        supervisorRuntimeScope: 'both',
+        supervisorNoProgressLimit: 2,
+        supervisorRepeatedToolLimit: 3,
+        reflectionEnabled: true,
+        reflectionAutoWriteMemory: true,
+        reflectionMinQuality: 0.8,
+      })
+
+      const result = await mod.observeAutonomyRunOutcome({
+        runId: 'run-low-quality',
+        sessionId: 's1',
+        agentId: 'agent-a',
+        source: 'chat',
+        status: 'completed',
+        resultText: 'I retried the same shell path and nothing changed.',
+        toolEvents: [
+          { name: 'shell', input: '{"cmd":"npm test"}' },
+          { name: 'shell', input: '{"cmd":"npm test"}' },
+          { name: 'shell', input: '{"cmd":"npm test"}' },
+        ],
+        mainLoopState: {
+          followupChainCount: 2,
+          summary: 'I retried the same shell path and nothing changed.',
+        },
+        sourceMessage: 'Repair the deployment workflow and keep notes for later.',
+      }, {
+        generateText: async () => JSON.stringify({
+          summary: 'Low quality reflection',
+          lessons: ['This weak note should not be written to memory.'],
+          quality_score: 0.25,
+          quality_reasoning: 'Too generic to keep as durable memory.',
+        }),
+      })
+
+      const memories = memoryMod.getMemoryDb().list(undefined, 50)
+        .filter((entry) => entry.metadata && entry.metadata.origin === 'autonomy-reflection')
+
+      console.log(JSON.stringify({
+        reflectionSummary: result.reflection?.summary ?? null,
+        qualityScore: result.reflection?.qualityScore ?? null,
+        autoMemoryCount: result.reflection?.autoMemoryIds?.length ?? 0,
+        storedReflectionMemoryCount: memories.length,
+      }))
+    `)
+
+    assert.equal(output.reflectionSummary, 'Low quality reflection')
+    assert.equal(output.qualityScore, 0.25)
+    assert.equal(output.autoMemoryCount, 0)
+    assert.equal(output.storedReflectionMemoryCount, 0)
+  })
+
+  it('skips semantically duplicate reflection memory when embedding dedup is enabled', () => {
+    const output = runWithTempDataDir(`
+      const http = await import('node:http')
+      const path = await import('node:path')
+      const Database = (await import('better-sqlite3')).default
+      const storageMod = await import('@/lib/server/storage')
+      const storage = storageMod.default || storageMod['module.exports'] || storageMod
+      const reflectionMod = await import('@/lib/server/autonomy/supervisor-reflection')
+      const mod = reflectionMod.default || reflectionMod['module.exports'] || reflectionMod
+      const memoryDbMod = await import('@/lib/server/memory/memory-db')
+      const memoryMod = memoryDbMod.default || memoryDbMod['module.exports'] || memoryDbMod
+      const settingsRepositoryMod = await import('@/lib/server/settings/settings-repository')
+      const settingsRepository = settingsRepositoryMod.default || settingsRepositoryMod['module.exports'] || settingsRepositoryMod
+
+      const server = http.createServer((req, res) => {
+        res.setHeader('content-type', 'application/json')
+        res.end(JSON.stringify({ embedding: [1, 0] }))
+      })
+      await new Promise((resolve) => server.listen(0, '127.0.0.1', resolve))
+      const endpoint = 'http://127.0.0.1:' + server.address().port
+
+      try {
+        storage.saveAgents({
+          'agent-a': {
+            id: 'agent-a',
+            name: 'Agent A',
+            provider: 'openai',
+            model: 'gpt-test',
+          },
+        })
+
+        storage.saveSessions({
+          s1: {
+            id: 's1',
+            name: 'Semantic Dedup Session',
+            cwd: process.cwd(),
+            user: 'tester',
+            provider: 'openai',
+            model: 'gpt-test',
+            claudeSessionId: null,
+            messages: [
+              { role: 'user', text: 'Release carefully and keep the durable lesson.', time: 1 },
+              { role: 'assistant', text: 'I checked the release gates before shipping.', time: 2 },
+            ],
+            createdAt: 1,
+            lastActiveAt: 2,
+            sessionType: 'human',
+            agentId: 'agent-a',
+          },
+        })
+
+        const memDb = memoryMod.getMemoryDb()
+        const existing = memDb.add({
+          agentId: 'agent-a',
+          sessionId: 's1',
+          category: 'reflection/lesson',
+          title: 'Reflection Lesson',
+          content: 'Always verify release gates before shipping.',
+          metadata: { origin: 'autonomy-reflection' },
+        })
+        const rawDb = new Database(path.join(process.env.DATA_DIR, 'memory.db'))
+        rawDb.prepare('UPDATE memories SET embedding = ? WHERE id = ?').run(Buffer.from(new Float32Array([1, 0]).buffer), existing.id)
+        rawDb.close()
+
+        settingsRepository.saveSettings({
+          supervisorEnabled: true,
+          supervisorRuntimeScope: 'both',
+          supervisorNoProgressLimit: 2,
+          supervisorRepeatedToolLimit: 3,
+          reflectionEnabled: true,
+          reflectionAutoWriteMemory: true,
+          reflectionSemanticDedupEnabled: true,
+          reflectionSemanticDedupThreshold: 0.9,
+          embeddingProvider: 'ollama',
+          embeddingModel: 'test-embedding',
+          embeddingEndpoint: endpoint,
+        })
+
+        const result = await mod.observeAutonomyRunOutcome({
+          runId: 'run-semantic-dedup',
+          sessionId: 's1',
+          agentId: 'agent-a',
+          source: 'chat',
+          status: 'completed',
+          resultText: 'I checked the release gates before shipping.',
+          toolEvents: [
+            { name: 'shell', input: '{"cmd":"npm test"}' },
+          ],
+          sourceMessage: 'Release carefully and keep the durable lesson.',
+        }, {
+          generateText: async () => JSON.stringify({
+            summary: 'Release gate reflection',
+            lessons: ['Confirm release gates before shipping.'],
+            quality_score: 0.95,
+          }),
+        })
+
+        const reflectionMemories = memDb.list('agent-a', 50)
+          .filter((entry) => entry.metadata && entry.metadata.origin === 'autonomy-reflection')
+
+        console.log(JSON.stringify({
+          reflectionSummary: result.reflection?.summary ?? null,
+          autoMemoryCount: result.reflection?.autoMemoryIds?.length ?? 0,
+          reflectionMemoryContents: reflectionMemories.map((entry) => entry.content).sort(),
+        }))
+      } finally {
+        await new Promise((resolve) => server.close(resolve))
+      }
+    `)
+
+    assert.equal(output.reflectionSummary, 'Release gate reflection')
+    assert.equal(output.autoMemoryCount, 0)
+    assert.deepEqual(output.reflectionMemoryContents, ['Always verify release gates before shipping.'])
+  })
+
   it('reflects short human chats when they contain durable personal context', () => {
     const output = runWithTempDataDir(`
       const storageMod = await import('@/lib/server/storage')