chore: pin ruff and ty linter versions

.pre-commit-config.yaml

@@ -9,17 +9,17 @@ repos:
     hooks:
       - id: ruff-format
         name: ruff format
-        entry: uvx ruff format --force-exclude --check
+        entry: uv run ruff format --force-exclude --check
         language: system
         types: [python]
       - id: ruff-check
         name: ruff check
-        entry: uvx ruff check --force-exclude --fix
+        entry: uv run ruff check --force-exclude --fix
         language: system
         types: [python]
       - id: ty-check
         name: ty type check
-        entry: uvx ty check .
+        entry: uv run ty check .
         language: system
         types: [python]
         pass_filenames: false

docs/DEVELOPMENT.md

@@ -34,7 +34,7 @@ Tip:
 ## Type Check
 
 ```bash
-uvx ty check .
+uv run ty check .
 ```
 
 ## Code Quality Checks
@@ -49,9 +49,12 @@ uv run ruff check .
 uv run ruff format --check .
 
 # Type check
-uvx ty check .
+uv run ty check .
 ```
 
+> **Linter versions are pinned.** `ruff` and `ty` are pinned to exact versions in the
+> `pyproject.toml` `[dependency-groups] dev` section (and locked in `uv.lock`) for consistent runs across local and CI.
+
 ## Pre-commit Hooks (Recommended)
 
 Automatically run quality checks before each commit using pre-commit hooks.

pyproject.toml

@@ -63,9 +63,9 @@ dev = [
     "pytest-asyncio>=1.0.0",
     "pytest-bdd>=7.2.0",
     "python-dotenv>=1.0.0",
-    "ty>=0.0.21",
+    "ty==0.0.64",
     "cryptography>=46.0.3",
-    "ruff>=0.8.0",
+    "ruff==0.16.0",
     "starlette>=0.40.0",
     "anyio>=3.6.2",
     "httpx>=0.27.0",

src/sap_cloud_sdk/adms/user-guide.md

@@ -45,6 +45,7 @@ relation = client.relations.create(
 
 # Upload bytes to the presigned URL (outside SDK)
 import requests
+
 upload_url = relation.document.document_content_upload_urls[0]
 requests.put(upload_url, data=open("Invoice.pdf", "rb"))
 ```
@@ -82,6 +83,7 @@ client = create_client(user_jwt=request.headers["Authorization"].split()[1])
 ```python
 from sap_cloud_sdk.adms import create_client, TokenCache
 
+
 # By default tokens are cached in-process via InMemoryTokenCache.
 # For multi-instance deployments (Kyma replicas > 1, CF instances > 1),
 # implement your own TokenCache subclass backed by the shared cache your
@@ -91,13 +93,21 @@ class MySharedCache(TokenCache):
     def set(self, key, token, ttl_seconds): ...
     def delete(self, key): ...
 
+
 client = create_client(token_cache=MySharedCache())
 ```
 
 ## Async Client
 
 ```python
-from sap_cloud_sdk.adms import create_async_client, BaseType, CreateDocumentInput, CreateDocumentRelationInput, RelationQueryOptions
+from sap_cloud_sdk.adms import (
+    create_async_client,
+    BaseType,
+    CreateDocumentInput,
+    CreateDocumentRelationInput,
+    RelationQueryOptions,
+)
+
 
 async def main():
     async with create_async_client() as client:
@@ -203,6 +213,7 @@ job = client.jobs.start_zip_download(params)
 
 # Poll until terminal state
 import time
+
 while not job.job_status or not job.job_status.is_terminal():
     time.sleep(2)
     job = client.jobs.get_status(job.job_id)

src/sap_cloud_sdk/agent_decorators/user-guide.md

@@ -21,6 +21,7 @@ from sap_cloud_sdk.agent_decorators import (
 ```python
 from sap_cloud_sdk.agent_decorators import prompt_section, agent_model
 
+
 # Define a prompt with a coded default
 @prompt_section(
     key="prompts.system",
@@ -30,6 +31,7 @@ from sap_cloud_sdk.agent_decorators import prompt_section, agent_model
 def system_prompt() -> str:
     return "You are a helpful assistant."
 
+
 # Define the model selection
 @agent_model(key="config.model", label="LLM Model")
 def model_name() -> str:
@@ -45,6 +47,7 @@ Expose a prompt section for editing.
 ```python
 from sap_cloud_sdk.agent_decorators import prompt_section
 
+
 @prompt_section(
     key="prompts.identity",
     label="Agent Identity",
@@ -62,6 +65,7 @@ Expose a configuration value for editing.
 ```python
 from sap_cloud_sdk.agent_decorators import agent_config
 
+
 @agent_config(
     key="config.temperature",
     label="Temperature",
@@ -78,6 +82,7 @@ Expose a model selection. The `description` parameter is optional.
 ```python
 from sap_cloud_sdk.agent_decorators import agent_model
 
+
 @agent_model(
     key="config.model",
     label="Default Model",
@@ -93,6 +98,7 @@ def default_model() -> str:
 from sap_cloud_sdk.agent_decorators.exceptions import AgentDecoratorError
 
 try:
+
     @prompt_section(key="", label="L", description="D")
     def bad():
         return ""

src/sap_cloud_sdk/agent_memory/user-guide.md

@@ -713,6 +713,7 @@ elif total <= PAGE_SIZE:
         agent_id="my-agent", invoker_id="user-123", limit=total
     )
 else:
+
     def iter_all_memories(client, agent_id, invoker_id, page_size=PAGE_SIZE):
         offset = 0
         while True:
@@ -780,7 +781,8 @@ The default `threshold` of `0.6` may be too strict for your data. Try a lower va
 
 ```python
 results = client.search_memories(
-    agent_id="my-agent", invoker_id="user-123",
+    agent_id="my-agent",
+    invoker_id="user-123",
     query="user display preferences",
     threshold=0.3,
 )
@@ -932,10 +934,10 @@ from sap_cloud_sdk.agent_memory.factory.langgraph_checkpoint import create_check
     key="config.checkpointer.ttl_seconds",
     label="Thread TTL (seconds)",
     description="Evict inactive conversation threads after this period of "
-                "inactivity. Set to 0 to disable eviction.",
+    "inactivity. Set to 0 to disable eviction.",
 )
 def thread_ttl_seconds() -> int:
-    return 3600 # 1 hour
+    return 3600  # 1 hour
 
 
 class MyAgent:

src/sap_cloud_sdk/agentgateway/user-guide.md

@@ -250,7 +250,7 @@ from sap_cloud_sdk.agentgateway import AgentCardFilter
 
 AgentCardFilter(
     agent_names=[],  # agent card names to include (matched against card JSON `name`); empty = no filter
-    ord_ids=[],      # ORD IDs to include (extracted from fragment URL); empty = no filter
+    ord_ids=[],  # ORD IDs to include (extracted from fragment URL); empty = no filter
 )
 ```
 
@@ -261,12 +261,14 @@ Both fields default to empty lists. Filters are applied with AND semantics: if b
 ```python
 @dataclass
 class Agent:
-    ord_id: str       # ORD ID from fragment ordId property
+    ord_id: str  # ORD ID from fragment ordId property
     agent_card: AgentCard
 
+
 @dataclass
 class AgentCard:
-    raw: dict         # full parsed JSON from /.well-known/agent-card.json
+    raw: dict  # full parsed JSON from /.well-known/agent-card.json
+
 
 @dataclass
 class MCPTool:

src/sap_cloud_sdk/aicore/user-guide.md

@@ -30,8 +30,7 @@ set_aicore_config()
 from litellm import completion
 
 response = completion(
-    model="sap/gpt-4",
-    messages=[{"role": "user", "content": "Hello!"}]
+    model="sap/gpt-4", messages=[{"role": "user", "content": "Hello!"}]
 )
 ```
 
@@ -115,25 +114,31 @@ from sap_cloud_sdk.aicore import (
     set_filtering,
 )
 
-set_filtering(ContentFiltering(
-    input_filtering=InputFiltering(filters=[
-        AzureContentFilter(
-            hate=Severity.STRICT,
-            violence=Severity.STRICT,
-            sexual=Severity.STRICT,
-            self_harm=Severity.STRICT,
-            prompt_shield=True,
+set_filtering(
+    ContentFiltering(
+        input_filtering=InputFiltering(
+            filters=[
+                AzureContentFilter(
+                    hate=Severity.STRICT,
+                    violence=Severity.STRICT,
+                    sexual=Severity.STRICT,
+                    self_harm=Severity.STRICT,
+                    prompt_shield=True,
+                ),
+            ]
         ),
-    ]),
-    output_filtering=OutputFiltering(filters=[
-        AzureContentFilter(
-            hate=Severity.MEDIUM,
-            violence=Severity.MEDIUM,
-            sexual=Severity.MEDIUM,
-            self_harm=Severity.MEDIUM,
+        output_filtering=OutputFiltering(
+            filters=[
+                AzureContentFilter(
+                    hate=Severity.MEDIUM,
+                    violence=Severity.MEDIUM,
+                    sexual=Severity.MEDIUM,
+                    self_harm=Severity.MEDIUM,
+                ),
+            ]
         ),
-    ]),
-))
+    )
+)
 ```
 
 To re-apply env-based config (e.g. after changing `AICORE_FILTER_*`):
@@ -161,12 +166,16 @@ from sap_cloud_sdk.aicore import (
     set_filtering,
 )
 
-set_filtering(ContentFiltering(
-    input_filtering=InputFiltering(filters=[
-        AzureContentFilter(prompt_shield=True),
-        LlamaGuard38bFilter(hate=True, violent_crimes=True),
-    ]),
-))
+set_filtering(
+    ContentFiltering(
+        input_filtering=InputFiltering(
+            filters=[
+                AzureContentFilter(prompt_shield=True),
+                LlamaGuard38bFilter(hate=True, violent_crimes=True),
+            ]
+        ),
+    )
+)
 ```
 
 `LlamaGuard38bFilter` takes 14 boolean category toggles (`hate`,
@@ -239,15 +248,23 @@ set_filtering(hate=0, violence=0)
 
 # After (aicore namespace with class API):
 from sap_cloud_sdk.aicore import (
-    AzureContentFilter, ContentFiltering, InputFiltering,
-    Severity, ContentFilteredError, set_filtering,
+    AzureContentFilter,
+    ContentFiltering,
+    InputFiltering,
+    Severity,
+    ContentFilteredError,
+    set_filtering,
 )
 
-set_filtering(ContentFiltering(
-    input_filtering=InputFiltering(filters=[
-        AzureContentFilter(hate=Severity.STRICT, violence=Severity.STRICT),
-    ]),
-))
+set_filtering(
+    ContentFiltering(
+        input_filtering=InputFiltering(
+            filters=[
+                AzureContentFilter(hate=Severity.STRICT, violence=Severity.STRICT),
+            ]
+        ),
+    )
+)
 ```
 
 Env vars also renamed: `ORCH_FILTER_*` → `AICORE_FILTER_*`. The
@@ -283,8 +300,8 @@ response = completion(
     model="sap/gpt-4",
     messages=[
         {"role": "system", "content": "You are a helpful assistant."},
-        {"role": "user", "content": "What is SAP AI Core?"}
-    ]
+        {"role": "user", "content": "What is SAP AI Core?"},
+    ],
 )
 
 print(response.choices[0].message.content)
@@ -301,7 +318,7 @@ set_aicore_config()
 response = completion(
     model="sap/gpt-4",
     messages=[{"role": "user", "content": "Tell me a story"}],
-    stream=True
+    stream=True,
 )
 
 for chunk in response:
@@ -319,14 +336,12 @@ set_aicore_config()
 
 # Chat completion
 chat_response = completion(
-    model="sap/gpt-4",
-    messages=[{"role": "user", "content": "Hello!"}]
+    model="sap/gpt-4", messages=[{"role": "user", "content": "Hello!"}]
 )
 
 # Text embeddings
 embedding_response = embedding(
-    model="sap/text-embedding-ada-002",
-    input=["Hello world", "How are you?"]
+    model="sap/text-embedding-ada-002", input=["Hello world", "How are you?"]
 )
 ```
 
@@ -356,10 +371,10 @@ try:
         model="sap/gpt-4",
         messages=[
             {"role": "system", "content": "You are a helpful AI assistant."},
-            {"role": "user", "content": "Explain SAP AI Core in one sentence."}
+            {"role": "user", "content": "Explain SAP AI Core in one sentence."},
         ],
         temperature=0.7,
-        max_tokens=100
+        max_tokens=100,
     )
 
     print(f"Response: {response.choices[0].message.content}")