Multi-runtime release-please + Python harness correctness

.github/workflows/release-please.yml

@@ -0,0 +1,31 @@
+name: release-please
+
+# Opens/maintains one "Release PR" per runtime and, when a Release PR is merged,
+# tags the runtime (e.g. python/2.0.0). Existing per-runtime publish workflows
+# (prompty-python.yml, prompty-rust.yml, prompty-ts-release.yml, prompty-csharp.yml)
+# trigger on those tags and do the actual package publishing.
+#
+# IMPORTANT: `token` must be a GitHub App installation token or a PAT — NOT the
+# default GITHUB_TOKEN. Tags created with GITHUB_TOKEN do not trigger the
+# tag-listening publish workflows (GitHub's recursive-workflow guard). Provide a
+# token via the RELEASE_PLEASE_TOKEN secret so the publish workflows fire.
+
+on:
+  push:
+    branches:
+      - main
+  workflow_dispatch:
+
+permissions:
+  contents: write
+  pull-requests: write
+
+jobs:
+  release-please:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: googleapis/release-please-action@a02a34c4d625f9be7cb89156071d8567266a2445 # v4
+        with:
+          token: ${{ secrets.RELEASE_PLEASE_TOKEN || github.token }}
+          config-file: release-please-config.json
+          manifest-file: .release-please-manifest.json

.release-please-manifest.json

@@ -0,0 +1,9 @@
+{
+  "runtime/python/prompty": "2.0.0-beta.3",
+  "runtime/rust": "2.0.0-beta.4",
+  "runtime/typescript/packages/core": "2.0.0-beta.5",
+  "runtime/csharp": "2.0.0-beta.4",
+  "runtime/java": "2.0.0-beta.4",
+  "runtime/go/prompty": "2.0.0-beta.4",
+  "runtime/swift/prompty": "2.0.0-beta.4"
+}

RELEASING.md

@@ -12,6 +12,81 @@ via OIDC (no secrets needed).
 | Python | [PyPI](https://pypi.org/project/prompty/) | `python/{version}` | `prompty-python.yml` |
 | TypeScript | [npm](https://www.npmjs.com/package/@prompty/core) | `typescript/{version}` | `prompty-ts-release.yml` |
 
+Those tags are now produced automatically by **release-please** (see below). The
+manual tag-and-push steps later in this document remain as a break-glass path.
+
+## Automated releases (release-please)
+
+Prompty ships **seven runtimes independently** via
+[release-please](https://github.com/googleapis/release-please). There is no shared
+version number — `python/2.1.0` and `rust/2.0.3` can coexist. Config lives in
+`release-please-config.json` + `.release-please-manifest.json` at the repo root, and
+the driver is `.github/workflows/release-please.yml`.
+
+> **VS Code extension** (`vscode/`) is intentionally **not** managed by
+> release-please — it uses the Marketplace pre-release channel (`vscode/2.0.0-pre.N`)
+> and keeps its own pipeline.
+
+### Flow
+
+```
+conventional commit on main
+        │
+        ▼
+release-please.yml  ── opens/updates one "Release PR" per runtime with changes
+        │              (title: "chore(<runtime>): release <ver>")
+        ▼
+merge the Release PR  ── bumps version files, writes CHANGELOG, pushes a tag
+        │                like `python/2.1.0`
+        ▼
+prompty-<runtime>.yml ── existing publish workflow triggers on the tag and publishes
+```
+
+release-please **only opens PRs and cuts tags**; nothing publishes until a human
+merges a Release PR. Commits route to a runtime by the **paths they touch**; add a
+scope when ambiguous (`feat(python):`, `fix(rust):`). Only `feat`, `fix`, `perf`,
+`revert`, `deps` trigger releases; `feat!:` / `BREAKING CHANGE:` bumps major.
+
+### Components
+
+| Component  | Path                               | release-type | Version source                                       | Publish |
+| ---------- | ---------------------------------- | ------------ | ---------------------------------------------------- | ------- |
+| `python`   | `runtime/python/prompty`           | `python`     | `prompty/_version.py` (`# x-release-please-version`) | ✅ PyPI  |
+| `rust`     | `runtime/rust`                     | `rust`       | all workspace `Cargo.toml` (native)                  | ✅ crates|
+| `typescript` | `runtime/typescript/packages/core` | `node`     | all `packages/*/package.json` (`node-workspace`)     | ✅ npm   |
+| `csharp`   | `runtime/csharp`                   | `simple`     | 4 `*.csproj` `<Version>` (annotated)                 | ✅ NuGet |
+| `java`     | `runtime/java`                     | `simple`     | `build.gradle.kts` `version` (annotated)             | check only |
+| `go`       | `runtime/go/prompty`               | `go`         | git tag only                                         | check only |
+| `swift`    | `runtime/swift/prompty`            | `simple`     | git tag only                                         | check only |
+
+Tag format (`separator:"/"`, `include-component-in-tag:true`, `include-v-in-tag:false`)
+reproduces the existing `python/2.1.0` tags exactly — **no publish workflow changed**.
+
+### Graduating beta → 2.0.0
+
+All runtimes were on `2.0.0-beta.N`. The manifest is seeded at each runtime's last
+beta with **no prerelease mode**, so the first Release PR graduates each runtime to a
+clean **`2.0.0`** stable release; normal `2.0.1` / `2.1.0` / `3.0.0` bumps follow.
+
+### One-time setup before enabling on `main`
+
+1. **`RELEASE_PLEASE_TOKEN` secret (required for auto-publish).** The workflow falls
+   back to `GITHUB_TOKEN`, but tags cut with `GITHUB_TOKEN` **do not trigger** the
+   tag-listening publish workflows. Provide a **GitHub App installation token** (or
+   fine-grained PAT) with `contents:write` + `pull_requests:write` as
+   `RELEASE_PLEASE_TOKEN`.
+2. **`bootstrap-sha`** in the config is pinned to the adoption commit so the first run
+   doesn't replay old history.
+
+### Verify on the first Release PR (before merging)
+
+- **typescript** — confirm all four `packages/*/package.json` versions **and** their
+  `^@prompty/core` ranges bump (`node-workspace`); if a sibling is missed, register it
+  as its own package in the config.
+- **rust** — confirm all workspace members + `path`/`version` deps bump together.
+- **csharp / java** — confirm the annotated `<Version>` / `version =` lines rewrite.
+- **python** — Release PR sets `_version.py` to `2.0.0`; the built wheel is `prompty-2.0.0`.
+
 ## Pre-flight checklist (MUST DO before every release)
 
 Run these **exact** steps locally before pushing tags. They mirror CI.

release-please-config.json

@@ -0,0 +1,66 @@
+{
+  "$schema": "https://raw.githubusercontent.com/googleapis/release-please/main/schemas/config.json",
+  "separator": "/",
+  "include-component-in-tag": true,
+  "include-v-in-tag": false,
+  "bootstrap-sha": "3c1c0cc27ee429983ee362aa60ac510678da4af9",
+  "changelog-sections": [
+    { "type": "feat", "section": "Features" },
+    { "type": "fix", "section": "Bug Fixes" },
+    { "type": "perf", "section": "Performance Improvements" },
+    { "type": "revert", "section": "Reverts" },
+    { "type": "deps", "section": "Dependencies" },
+    { "type": "docs", "section": "Documentation", "hidden": true },
+    { "type": "chore", "section": "Miscellaneous", "hidden": true },
+    { "type": "ci", "section": "Continuous Integration", "hidden": true },
+    { "type": "test", "section": "Tests", "hidden": true },
+    { "type": "refactor", "section": "Code Refactoring", "hidden": true }
+  ],
+  "packages": {
+    "runtime/python/prompty": {
+      "component": "python",
+      "release-type": "python",
+      "package-name": "prompty",
+      "extra-files": [
+        { "type": "generic", "path": "prompty/_version.py" }
+      ]
+    },
+    "runtime/rust": {
+      "component": "rust",
+      "release-type": "rust"
+    },
+    "runtime/typescript/packages/core": {
+      "component": "typescript",
+      "release-type": "node",
+      "package-name": "@prompty/core"
+    },
+    "runtime/csharp": {
+      "component": "csharp",
+      "release-type": "simple",
+      "extra-files": [
+        { "type": "generic", "path": "Prompty.Core/Prompty.Core.csproj" },
+        { "type": "generic", "path": "Prompty.OpenAI/Prompty.OpenAI.csproj" },
+        { "type": "generic", "path": "Prompty.Foundry/Prompty.Foundry.csproj" },
+        { "type": "generic", "path": "Prompty.Anthropic/Prompty.Anthropic.csproj" }
+      ]
+    },
+    "runtime/java": {
+      "component": "java",
+      "release-type": "simple",
+      "extra-files": [
+        { "type": "generic", "path": "build.gradle.kts" }
+      ]
+    },
+    "runtime/go/prompty": {
+      "component": "go",
+      "release-type": "go"
+    },
+    "runtime/swift/prompty": {
+      "component": "swift",
+      "release-type": "simple"
+    }
+  },
+  "plugins": [
+    "node-workspace"
+  ]
+}

runtime/csharp/Prompty.Anthropic/Prompty.Anthropic.csproj

@@ -4,7 +4,7 @@
     <TargetFramework>net9.0</TargetFramework>
     <ImplicitUsings>enable</ImplicitUsings>
     <Nullable>enable</Nullable>
-    <Version>2.0.0-beta.4</Version>
+    <Version>2.0.0-beta.4</Version> <!-- x-release-please-version -->
     <PackageId>Prompty.Anthropic</PackageId>
     <Authors>Microsoft</Authors>
     <Description>Anthropic provider for Prompty — executor and processor for Claude models via the Anthropic Messages API.</Description>

runtime/csharp/Prompty.Core/Prompty.Core.csproj

@@ -4,7 +4,7 @@
     <TargetFramework>net9.0</TargetFramework>
     <ImplicitUsings>enable</ImplicitUsings>
     <Nullable>enable</Nullable>
-    <Version>2.0.0-beta.4</Version>
+    <Version>2.0.0-beta.4</Version> <!-- x-release-please-version -->
     <PackageId>Prompty.Core</PackageId>
     <Authors>Microsoft</Authors>
     <Description>Prompty is an asset class and format for LLM prompts. Core library with loader, pipeline, renderers, parsers, and tracing.</Description>

runtime/csharp/Prompty.Foundry/Prompty.Foundry.csproj

@@ -6,7 +6,7 @@
     <Nullable>enable</Nullable>
     <NoWarn>OPENAI001;CS1591</NoWarn>
     <GenerateDocumentationFile>true</GenerateDocumentationFile>
-    <Version>2.0.0-beta.4</Version>
+    <Version>2.0.0-beta.4</Version> <!-- x-release-please-version -->
     <PackageId>Prompty.Foundry</PackageId>
     <Authors>Microsoft</Authors>
     <Description>Microsoft Foundry (Azure OpenAI) provider for Prompty — executor and processor for Azure-hosted models.</Description>

runtime/csharp/Prompty.OpenAI/Prompty.OpenAI.csproj

@@ -6,7 +6,7 @@
     <Nullable>enable</Nullable>
     <NoWarn>OPENAI001;CS1591</NoWarn>
     <GenerateDocumentationFile>true</GenerateDocumentationFile>
-    <Version>2.0.0-beta.4</Version>
+    <Version>2.0.0-beta.4</Version> <!-- x-release-please-version -->
     <PackageId>Prompty.OpenAI</PackageId>
     <Authors>Microsoft</Authors>
     <Description>OpenAI provider for Prompty — executor and processor for OpenAI chat, embedding, image, and agent APIs.</Description>

runtime/java/build.gradle.kts

@@ -4,7 +4,7 @@ plugins {
 
 allprojects {
     group = "com.microsoft.prompty"
-    version = "2.0.0-beta.4"
+    version = "2.0.0-beta.4" // x-release-please-version
 }
 
 subprojects {

runtime/python/prompty/.env.example

@@ -9,7 +9,7 @@ OPENAI_API_KEY=
 OPENAI_BASE_URL=
 OPENAI_MODEL=gpt-4o-mini
 OPENAI_EMBEDDING_MODEL=text-embedding-3-small
-OPENAI_IMAGE_MODEL=dall-e-2
+OPENAI_IMAGE_MODEL=gpt-image-1
 
 # Direct OpenAI (api.openai.com — no proxy/compat layer)
 DIRECT_OPENAI_API_KEY=

runtime/python/prompty/prompty/__init__.py

@@ -139,8 +139,6 @@
     "bind_tools",
     "tool",
     # Backward-compat aliases
-    "AzureExecutor",
-    "AzureProcessor",
     "PromptAgent",
     "AgentDefinition",
     # Model discovery
@@ -280,7 +278,5 @@
 )
 
 # Backward-compat aliases (will be removed in a future version)
-AzureExecutor = FoundryExecutor
-AzureProcessor = FoundryProcessor
 PromptAgent = Agent
 AgentDefinition = Agent

runtime/python/prompty/prompty/_version.py

@@ -1 +1 @@
-VERSION = "2.0.0b3"
+VERSION = "2.0.0-beta.3"  # x-release-please-version

runtime/python/prompty/prompty/core/discovery.py

@@ -172,11 +172,11 @@ def _discover(group: str, key: str) -> Any:
             return _cache[cache_key]
 
     eps = importlib.metadata.entry_points(group=group, name=key)
-    ep_list = list(eps) if not isinstance(eps, list) else eps
-    if not ep_list:
+    first = next(iter(eps), None)
+    if first is None:
         raise InvokerError(group, key)
 
-    loaded = _instantiate(ep_list[0].load())
+    loaded = _instantiate(first.load())
     with _lock:
         _cache[cache_key] = loaded
     return loaded

runtime/python/prompty/pyproject.toml

@@ -96,7 +96,7 @@ ignore = ["E501"]
 [tool.pytest.ini_options]
 asyncio_mode = "strict"
 markers = [
-  "integration: tests that hit real OpenAI / Azure OpenAI endpoints (deselected by default)",
+  "integration: tests that hit real OpenAI / Azure OpenAI (Foundry) endpoints (deselected by default)",
 ]
 addopts = "-m 'not integration'"
 

runtime/python/prompty/tests/integration/conftest.py

@@ -17,7 +17,10 @@
 
 from __future__ import annotations
 
+import json
 import os
+from functools import cache
+from pathlib import Path
 from typing import Any
 
 import pytest
@@ -46,7 +49,7 @@ def pytest_collection_modifyitems(items: list[pytest.Item]) -> None:
 _OPENAI_BASE_URL = os.environ.get("OPENAI_BASE_URL", "")  # optional: proxy via Azure
 _OPENAI_MODEL = os.environ.get("OPENAI_MODEL", "gpt-4o-mini")  # override default chat model
 _OPENAI_EMBEDDING_MODEL = os.environ.get("OPENAI_EMBEDDING_MODEL", "text-embedding-3-small")
-_OPENAI_IMAGE_MODEL = os.environ.get("OPENAI_IMAGE_MODEL", "dall-e-2")
+_OPENAI_IMAGE_MODEL = os.environ.get("OPENAI_IMAGE_MODEL", "gpt-image-1")
 _AZURE_KEY = os.environ.get("AZURE_OPENAI_API_KEY", "")
 _AZURE_ENDPOINT = os.environ.get("AZURE_OPENAI_ENDPOINT", "")
 _AZURE_CHAT_DEPLOYMENT = os.environ.get("AZURE_OPENAI_CHAT_DEPLOYMENT", "")
@@ -405,3 +408,60 @@ def make_entra_agent(
     if metadata is not None:
         data["metadata"] = metadata
     return Agent.load(data)
+
+
+# ---------------------------------------------------------------------------
+# Conformance vector access
+# ---------------------------------------------------------------------------
+#
+# Live agent-loop tests are anchored to the SAME canonical @vector corpus the
+# deterministic model-conformance suite replays (schema/tsp-output/
+# .typra-generated/vectors.json, emitted by Typra from the TypeSpec sources).
+# The live test reuses a vector's scenario — its messages, tools, and derived
+# tool-dispatch order — but swaps the scripted model double for a real provider,
+# so one Typra-governed definition validates both the deterministic engine and
+# the live provider path.
+
+# repo root: tests/integration -> tests -> prompty -> python -> runtime -> root
+_VECTORS_JSON = Path(__file__).resolve().parents[5] / "schema" / "tsp-output" / ".typra-generated" / "vectors.json"
+
+
+@cache
+def _load_vectors_corpus() -> tuple[dict[str, Any], ...]:
+    if not _VECTORS_JSON.exists():
+        raise FileNotFoundError(
+            f"Conformance vector corpus not found at {_VECTORS_JSON}. "
+            "Regenerate it via the Typra emitter (schema build)."
+        )
+    with open(_VECTORS_JSON, encoding="utf-8") as handle:
+        return tuple(json.load(handle).get("vectors", []))
+
+
+def load_turn_vector(name: str) -> dict[str, Any]:
+    """Return the ``TurnConformance.run`` vector named ``name`` from the corpus.
+
+    Raises ``KeyError`` if no such vector exists so a renamed or removed vector
+    fails loudly rather than silently skipping live coverage.
+    """
+    for entry in _load_vectors_corpus():
+        if (
+            entry.get("contract") == "TurnConformance"
+            and entry.get("operation") == "run"
+            and isinstance(entry.get("vector"), dict)
+            and entry["vector"].get("name") == name
+        ):
+            return entry["vector"]
+    raise KeyError(f"TurnConformance.run vector '{name}' not found in {_VECTORS_JSON}")
+
+
+def turn_vector_tool_order(vector: dict[str, Any]) -> list[str]:
+    """Derive the expected ordered tool-dispatch sequence from a run vector.
+
+    Reads each scripted turn's ``expected_tool_calls`` in order — this is the
+    dependency chain the live provider must reproduce.
+    """
+    order: list[str] = []
+    for step in vector.get("sequence", []):
+        for call in step.get("expected_tool_calls") or []:
+            order.append(call["name"])
+    return order