burgan-tech · yilmaztayfun · Jul 27, 2026 · Jul 27, 2026
diff --git a/blog/2026-07-27-v0-0-76.md b/blog/2026-07-27-v0-0-76.md
@@ -0,0 +1,154 @@
+---
+slug: release-v0-0-76
+title: Release v0.0.76
+authors: [vnext-team]
+tags: [release, duyuru, tasks, caching, ai]
+date: 2026-07-27
+---
+
+## Overview
+
+This release adds AI to the task palette and sharpens the caching and concurrency primitives. A new **Dapr Conversation task** (`TaskType = 20`) invokes an LLM/AI provider (OpenAI, Anthropic, Bedrock, …) provider-agnostically through the Dapr Conversation building block ([#844](https://github.com/burgan-tech/vnext/issues/844)). **Function result caching gains vary-by headers** — `varyByHeaders` / `varyByHeaderPrefixes` feed a `varyKey(context)` key-expression helper so one function keeps separate cache variants per header value ([#839](https://github.com/burgan-tech/vnext/issues/839)). Flow-level cache tuning is **consolidated under a single `attributes.config` object** (`config.functionCache.ttlSeconds`), the author-controlled TTL for the built-in instance functions ([#846](https://github.com/burgan-tech/vnext/issues/846)). **Distributed resource locks** become production-ready with **idempotent release and automatic terminal cleanup** ([#840](https://github.com/burgan-tech/vnext/issues/840)), plus a per-subInstance lock key for subflow terminal-outcome propagation ([#845](https://github.com/burgan-tech/vnext/issues/845)). **Wizard states are now treated like normal states** in the State/View functions ([#838](https://github.com/burgan-tech/vnext/issues/838)). The release also fixes author `Content-Type` preservation ([#847](https://github.com/burgan-tech/vnext/issues/847)), case-insensitive request-header normalization ([#841](https://github.com/burgan-tech/vnext/issues/841)), and init mappings support for publish ([#848](https://github.com/burgan-tech/vnext/issues/848)). This release runs on component **schema 0.0.51**.
+
+{/* truncate */}
+
+---
+
+## Features
+
+### Dapr Conversation task — provider-agnostic LLM calls (#844)
+
+A new task type, **DaprConversation** (`TaskType = 20`), invokes an LLM/AI provider through Dapr's Conversation building block — the same sidecar the other Dapr task types use. The provider and its credentials live in a **domain-owned** Dapr `conversation` component (e.g. `openai`); the task only references the component name and the messages.
+
+```json
+{
+  "attributes": {
+    "type": "20",
+    "config": {
+      "componentName": "openai",
+      "inputs": [
+        { "role": "system", "content": "You summarize customer complaints." },
+        { "role": "user", "content": "Summarize the complaint in two sentences." }
+      ],
+      "parameters": { "model": "gpt-4o-mini", "maxTokens": "512" },
+      "temperature": 0.2,
+      "scrubPII": true
+    }
+  }
+}
+```
+
+- Required config is just `componentName`; `inputs` is an array of `role`/`content` messages, and `parameters` carries provider-specific string values (`model`, `maxTokens`, …).
+- Optional `metadata`, `contextId` (stateful conversations), `temperature`, `scrubPII`, and `timeoutSeconds` (default 30).
+- The runtime component is installed via the helm chart ([vnext-helm-charts #28](https://github.com/burgan-tech/vnext-helm-charts/pull/28)), so shipping a new provider needs no runtime redeploy.
+
+> **Reference:** issue [#844](https://github.com/burgan-tech/vnext/issues/844), helm PR [vnext-helm-charts #28](https://github.com/burgan-tech/vnext-helm-charts/pull/28) — see also the new [Dapr Conversation Task](/docs/components/tasks/dapr-conversation) component page.
+
+### Function cache vary-by headers (#839)
+
+The function `attributes.cache` block gains two fields — **`varyByHeaders`** (exact request-header names) and **`varyByHeaderPrefixes`** (header-name prefixes) — that fold the named headers into the cache key. They form the header-name set for the `varyKey(context)` key-expression helper (the two lists are unioned), so a single read function keeps **separate cache variants per header value** without the runtime baking in any header convention. When the instance does not supply its own `Instance.Data["varyBy"]`, these domain-declared fields are used.
+
+> **Reference:** issue [#839](https://github.com/burgan-tech/vnext/issues/839) — see also [Custom Functions → Cache](/docs/components/functions/custom).
+
+### Flow-level cache config object (#846)
+
+Author-controlled, flow-scoped cache tuning is now consolidated under a single **`attributes.config`** object. Its first member, **`functionCache.ttlSeconds`**, is the TTL for this workflow's built-in instance functions (`data`, `view`, `schema`, …).
+
+```json
+{ "config": { "functionCache": { "ttlSeconds": 120 } } }
+```
+
+- Repeat calls for the **same instance** are served from cache for the TTL window; when the **instance changes, the cache drops** and the next call is re-cached.
+- The host default is **60s** when `config` (or `ttlSeconds`) is absent or non-positive.
+- The **State Function is excluded** — the platform manages its cache separately (host-side `StateFunctionCache`).
+
+> **Reference:** issue [#846](https://github.com/burgan-tech/vnext/issues/846) — see also [Workflow component → Config](/docs/components/workflow).
+
+### Distributed resource locks — idempotent release + automatic cleanup (#840, #845)
+
+Start, state-level, and shared transitions can declare an optional **`resourceLock`** block — a distributed lock (Dapr `lock.redis`, via the Aether SDK) that prevents concurrent instances from mutating a shared resource (a seat, daily limit, account, …). It runs in the **Manual** profile only, is owned by the `instanceId`, and always carries a TTL.
+
+```json
+{
+  "key": "reserve-daily-limit",
+  "target": "limit-reserved",
+  "triggerType": 0,
+  "versionStrategy": "Patch",
+  "resourceLock": {
+    "keyExpression": { "location": "./src/DailyLimitKey.csx", "code": "<base64>", "type": "L", "encoding": "NAT" },
+    "action": "Acquire",
+    "ttlSeconds": 300,
+    "onConflict": "Abort"
+  }
+}
+```
+
+- The recommended model is **`Acquire` on the check/entry transition** and let the runtime release it: when the instance reaches a terminal state (Completed / Faulted / Cancelled), its locks are **released automatically** — no manual `Release` on every terminal transition.
+- **`Release` is idempotent/best-effort** — `LockDoesNotExist` counts as success, a foreign lock is logged (not faulted), and lock cleanup never rolls back a successful business transition.
+- A conflict on `Acquire` aborts the transition and returns **HTTP 409** (the instance is marked faulted `F`); the caller retries. `Extend` is unreliable (no native Dapr extend) — size the TTL to cover the whole operation.
+- Subflows now derive a **per-subInstance lock key** so terminal-outcome propagation stays correct across nested instances ([#845](https://github.com/burgan-tech/vnext/issues/845)).
+
+> **Reference:** issues [#840](https://github.com/burgan-tech/vnext/issues/840), [#845](https://github.com/burgan-tech/vnext/issues/845) — see also the new [Resource Lock guide](/docs/how-to/resource-lock) and [Workflow component → Transition](/docs/components/workflow).
+
+### Wizard states behave like normal states in read functions (#838)
+
+The State and View functions now treat **Wizard states like normal states**: authorization/role evaluation and the available-transition list are resolved the same way, so a wizard state's `state`/`view` response is consistent with any other state. This removes the special-casing that previously diverged wizard-state reads.
+
+> **Reference:** issue [#838](https://github.com/burgan-tech/vnext/issues/838) — see also [Workflow component → Wizard State ve View Davranışı](/docs/components/workflow).
+
+---
+
+## Fixes
+
+- **Author `Content-Type` preserved without charset suffix** — function output scripts that set `content-type` keep the author value verbatim (the `; charset=...` suffix is no longer appended) ([#847](https://github.com/burgan-tech/vnext/issues/847)).
+- **Case-insensitive request-header normalization** — transition request headers are matched case-insensitively, so header lookups no longer miss on casing differences ([#841](https://github.com/burgan-tech/vnext/issues/841)).
+- **Init mappings support for publish** — the init/publish path handles mapping components correctly ([#848](https://github.com/burgan-tech/vnext/issues/848)).
+
+---
+
+## Configuration Updates
+
+Configuration for v0.0.76:
+
+```json
+{
+  "runtimeVersion": "0.0.76",
+  "schemaVersion": "0.0.51"
+}
+```
+
+> **Note:** Schema version advances to **0.0.51** (from `0.0.50`). The bump adds the DaprConversation task type `"20"` ([#844](https://github.com/burgan-tech/vnext/issues/844)), the function-cache `varyByHeaders` / `varyByHeaderPrefixes` fields ([#839](https://github.com/burgan-tech/vnext/issues/839)), the workflow-level `attributes.config` object (`config.functionCache`) ([#846](https://github.com/burgan-tech/vnext/issues/846)), and the transition `resourceLock` definition ([#840](https://github.com/burgan-tech/vnext/issues/840)). Update `@burgan-tech/vnext-schema` in your domain project before validating against this runtime.
+
+**Container images:** published at tag `0.0.76` under `ghcr.io/burgan-tech/vnext/*`, Cosign-signed (keyless OIDC) with SBOM + provenance. Immutable digests are listed in the [GitHub release](https://github.com/burgan-tech/vnext/releases/tag/v0.0.76).
+
+---
+
+## Issues Referenced
+
+- [vnext #844](https://github.com/burgan-tech/vnext/issues/844) — Dapr Conversation (AI/LLM) task type.
+- [vnext #839](https://github.com/burgan-tech/vnext/issues/839) — `varyKey` cache-key helper + config-driven vary-by headers.
+- [vnext #846](https://github.com/burgan-tech/vnext/issues/846) — Consolidate flow-level `functionCache` under a `config` object.
+- [vnext #840](https://github.com/burgan-tech/vnext/issues/840) — Resource lock: idempotent release + automatic terminal cleanup.
+- [vnext #845](https://github.com/burgan-tech/vnext/issues/845) — Retryable 503 + per-subInstance lock key for terminal-outcome propagation.
+- [vnext #838](https://github.com/burgan-tech/vnext/issues/838) — Treat wizard states like normal states in State/View functions.
+- [vnext #847](https://github.com/burgan-tech/vnext/issues/847) — Preserve author `Content-Type` without charset suffix.
+- [vnext #841](https://github.com/burgan-tech/vnext/issues/841) — Normalize request headers case-insensitively.
+- [vnext #848](https://github.com/burgan-tech/vnext/issues/848) — Init mappings support for publish.
+- [vnext-helm-charts #28](https://github.com/burgan-tech/vnext-helm-charts/pull/28) — Dapr Conversation helm component.
+
+---
+
+## Summary
+
+- **Dapr Conversation task (Type 20)** invokes an LLM/AI provider provider-agnostically through the Dapr Conversation building block; the provider is a domain-owned Dapr component.
+- **Function cache vary-by headers**: `varyByHeaders` / `varyByHeaderPrefixes` feed the `varyKey(context)` helper for per-header cache variants.
+- **Flow-level `config.functionCache.ttlSeconds`**: author-controlled TTL for built-in instance functions (host default 60s; State Function excluded).
+- **Distributed resource locks** are production-ready: idempotent release, automatic terminal cleanup, HTTP 409 on conflict, and a per-subInstance lock key for subflows.
+- **Wizard states** are read like normal states in the State/View functions.
+- **Fixes**: author `Content-Type` preserved, case-insensitive header matching, init mappings support for publish.
+- **Schema** is **0.0.51**.
+
+---
+
+**vNext Runtime Platform Team**
+Released July 27, 2026
diff --git a/docs/components/functions/custom.md b/docs/components/functions/custom.md
@@ -251,6 +251,12 @@ Cache, fonksiyon başına **opt-in**'dir ve yalnızca **yan etkisiz (read) fonks
 | `bypassOnCacheError` | `boolean` | Hayır | `true` | `true`: cache okuma/yazma hataları isteği bozmaz, fonksiyon normal çalıştırılır. `false`: cache hatası isteği başarısız kılar |
 | `generationKeyExpression` | `object` | Hayır | — | Generation stamp'inin tutulduğu state key'ini çözen Dynamic Expresso ifadesi. `generationKey`'den önceliklidir |
 | `generationKey` | `string` | Hayır | — | Generation stamp'ini tutan statik state key'i |
+| `varyByHeaders` | `string[]` | Hayır | — | Sonucu değiştiren **tam (exact)** request-header adları. Belirtilen header'lar cache key'ine katılır |
+| `varyByHeaderPrefixes` | `string[]` | Hayır | — | Sonucu değiştiren request-header adı **prefiksleri** (ör. bir prefix ile başlayan tüm header'lar) |
+
+:::tip Vary-by header'lar (`#839`)
+`varyByHeaders` ve `varyByHeaderPrefixes`, key-expression içinde kullanılabilen `varyKey(context)` yardımcı fonksiyonunun header-adı kümesini oluşturur (ikisi birleştirilir). Böylece runtime herhangi bir header konvansiyonuna bağlı kalmadan; aynı fonksiyonun farklı header değerleri için **ayrı cache varyantları** tutar. Instance `Instance.Data["varyBy"]` ile kendi kümesini sağlamazsa, domain'de tanımlı bu iki alan kullanılır.
+:::
 
 ### Generation-Namespace Invalidation
 

diff --git a/docs/components/tasks/dapr-conversation.md b/docs/components/tasks/dapr-conversation.md
@@ -0,0 +1,141 @@
+---
+sidebar_position: 15
+title: Dapr Conversation Task
+description: Dapr Conversation building block üzerinden LLM/AI sağlayıcı çağrısı yapan task
+---
+
+# Dapr Conversation Task (Type: `20`)
+
+Dapr Conversation Task (`type: "20"`), bir **LLM/AI sağlayıcısını** Dapr'ın **Conversation building block**'u üzerinden çağırır. Böylece iş akışları; OpenAI, Anthropic, AWS Bedrock gibi sağlayıcılara **sağlayıcı-bağımsız** tek bir arayüzle prompt gönderebilir — diğer Dapr task türlerinin kullandığı aynı Dapr sidecar üzerinden.
+
+Sağlayıcı seçimi ve kimlik bilgileri **domain'e ait** bir Dapr `conversation` component'inde (ör. `openai`) tanımlanır; task yalnızca o component'in adını (`componentName`) ve mesajları (`inputs`) referans verir. Runtime bileşenini eklemek için bkz. [vnext-helm-charts #28](https://github.com/burgan-tech/vnext-helm-charts/pull/28).
+
+## Görev Tanımı
+
+> **Schema:** `task-definition.schema.json`
+
+```json
+{
+  "key": "summarize-complaint",
+  "version": "1.0.0",
+  "domain": "core",
+  "flow": "sys-tasks",
+  "flowVersion": "1.0.0",
+  "tags": ["ai", "llm", "conversation"],
+  "attributes": {
+    "type": "20",
+    "config": {
+      "componentName": "openai",
+      "inputs": [
+        { "role": "system", "content": "Sen bir müşteri şikayeti özetleyicisisin." },
+        { "role": "user", "content": "Şikayet metnini 2 cümlede özetle." }
+      ],
+      "parameters": {
+        "model": "gpt-4o-mini",
+        "maxTokens": "512"
+      },
+      "temperature": 0.2,
+      "scrubPII": true,
+      "timeoutSeconds": 30
+    }
+  }
+}
+```
+
+## Konfigürasyon Alanları
+
+| Alan | Tip | Zorunlu | Varsayılan | Açıklama |
+|------|-----|---------|------------|----------|
+| `componentName` | string | **Evet** | — | Dapr conversation component adı (yapılandırılmış LLM sağlayıcısı), ör. `openai` |
+| `inputs` | array | Hayır | — | Konuşma girdileri: `role`/`content` mesajlarından oluşan dizi (aşağıya bakın) |
+| `parameters` | object | Hayır | — | Sağlayıcıya özgü **string** parametreler (ör. `model`, `maxTokens`). Component'e olduğu gibi iletilir |
+| `metadata` | object | Hayır | — | İstekle iletilen Dapr component metadata'sı (string değerler) |
+| `contextId` | string | Hayır | — | Durumlu (stateful) bir konuşmayı sürdürmek için bağlam tanımlayıcısı |
+| `temperature` | number | Hayır | — | Örnekleme sıcaklığı (sampling temperature) |
+| `scrubPII` | boolean | Hayır | — | `true` ise sağlayıcıdan prompt ve yanıtlarda PII temizliği (scrub) istenir |
+| `timeoutSeconds` | integer | Hayır | `30` | Timeout süresi (saniye, minimum: 1) |
+
+### `inputs` Mesaj Yapısı
+
+Her girdi bir rol/içerik mesajıdır:
+
+```json
+{ "role": "user", "content": "...", "scrubPII": true, "name": "opsiyonel" }
+```
+
+| Alan | Değerler | Açıklama |
+|------|----------|----------|
+| `role` | `user`, `system`, `assistant`, `developer`, `tool` | Mesajın rolü |
+| `content` | string | Mesaj içeriği |
+| `scrubPII` | boolean (ops.) | Bu mesaj için PII temizliği |
+| `name` | string (ops.) | Mesaj için opsiyonel ad |
+
+## Property Erişimi
+
+Değerler çoğunlukla statik konfigürasyon yerine input mapping içinde dinamik olarak atanır.
+
+| Property | Setter Metodu | Açıklama |
+|----------|---------------|----------|
+| `ComponentName` | `SetComponentName(string componentName)` | Conversation component adı |
+| `Inputs` | `SetInputs(dynamic inputs)` | Mesaj dizisi |
+| `Parameters` | `SetParameters(Dictionary<string, string?> parameters)` | Sağlayıcı parametreleri |
+| `Metadata` | `SetMetadata(Dictionary<string, string?> metadata)` | Component metadata'sı |
+| `ContextId` | `SetContextId(string? contextId)` | Bağlam tanımlayıcısı |
+| `Temperature` | `SetTemperature(double? temperature)` | Örnekleme sıcaklığı |
+| `ScrubPII` | `SetScrubPII(bool? scrubPII)` | PII temizliği |
+| `TimeoutSeconds` | `SetTimeoutSeconds(int? timeoutSeconds)` | Timeout süresi |
+
+Input mapping örneği:
+
+```csharp
+public class SummarizeComplaintMapping : ScriptBase, IMapping
+{
+    public Task<ScriptResponse> InputHandler(WorkflowTask task, ScriptContext context)
+    {
+        var conversation = task as DaprConversationTask;
+        conversation.SetComponentName("openai");
+        conversation.SetInputs(new[]
+        {
+            new { role = "system", content = "Sen bir müşteri şikayeti özetleyicisisin." },
+            new { role = "user", content = (string)context.Instance.Data.complaintText }
+        });
+        conversation.SetParameters(new Dictionary<string, string?>
+        {
+            ["model"] = "gpt-4o-mini",
+            ["maxTokens"] = "512"
+        });
+        return Task.FromResult(new ScriptResponse());
+    }
+
+    public Task<ScriptResponse> OutputHandler(ScriptContext context)
+    {
+        var summary = context.Body?.outputs?[0]?.result;
+        return Task.FromResult(new ScriptResponse { Data = new { summary } });
+    }
+}
+```
+
+## Standart Yanıt
+
+Yanıt, sağlayıcının döndürdüğü çıktı(lar)ı `outputs` altında taşır:
+
+```json
+{
+  "Data": {
+    "outputs": [
+      { "result": "Müşteri, kartının teslim edilmediğini ve çağrı merkezine ulaşamadığını bildiriyor." }
+    ]
+  },
+  "StatusCode": 200,
+  "IsSuccess": true,
+  "ErrorMessage": null,
+  "TaskType": "20"
+}
+```
+
+## İlgili
+
+- [Tasks Genel Bakış](/docs/components/tasks/) — tüm task türleri ve referans mekanizması
+- [DaprService Task](/docs/components/tasks/dapr-service) — aynı Dapr sidecar üzerinden service invocation
+- Schema kaynağı: [task-definition.schema.json (vnext-schema)](https://github.com/burgan-tech/vnext-schema/blob/master/schemas/task-definition.schema.json)
+- Helm component: [vnext-helm-charts #28](https://github.com/burgan-tech/vnext-helm-charts/pull/28)
diff --git a/docs/components/tasks/index.md b/docs/components/tasks/index.md
@@ -53,7 +53,7 @@ Her task tanımı `task-definition.schema.json` şemasına uyar. Zorunlu alanlar
 
 ## Görev Türleri
 
-`task-definition.schema.json` toplamda **19 task türü** tanımlar:
+`task-definition.schema.json` toplamda **20 task türü** tanımlar:
 
 | # | Görev Türü | Açıklama | Detay |
 |---|---|---|---|
@@ -76,6 +76,7 @@ Her task tanımı `task-definition.schema.json` şemasına uyar. Zorunlu alanlar
 | 17 | **StateStoreTask** | Dapr state store ile cache (get/set/delete) | [StateStore](./state-store) |
 | 18 | **CacheAsideTask** | Read-through cache (miss'te sourceTask çalıştırıp cache'ler) | [CacheAside](./cache-aside) |
 | 19 | **GetInstanceTask** | Tek bir instance'ın tam projeksiyonunu (metadata + data) çekme | [GetInstance](./get-instance) |
+| 20 | **DaprConversationTask** | Dapr Conversation ile LLM/AI sağlayıcı çağrısı | [DaprConversation](./dapr-conversation) |
 
 ## Görev Kullanımı