Bump version to 0.7.2

Author: asimurka

docs/devel_doc/openapi.json

@@ -258,7 +258,7 @@ make run CONFIG=examples/lightspeed-stack-azure-entraid-service.yaml
 
 | Name                   | Type   | Pip Dependencies                                                                                     | Supported in LCS |
 |------------------------|--------|------------------------------------------------------------------------------------------------------|:----------------:|
-| rag-runtime            | inline | `chardet`,`pypdf`, `tqdm`, `numpy`, `scikit-learn`, `scipy`, `nltk`, `sentencepiece`, `transformers` | ✅               |
+| file-search            | inline | `chardet`,`pypdf`, `tqdm`, `numpy`, `scikit-learn`, `scipy`, `nltk`, `sentencepiece`, `transformers` | ✅               |
 | bing-search            | remote | `requests`                                                                                           | ❌               |
 | brave-search           | remote | `requests`                                                                                           | ❌               |
 | model-context-protocol | remote | `mcp>=1.8.1`                                                                                         | ✅               |

docs/devel_doc/providers.md

@@ -2,7 +2,7 @@ version: 2
 image_name: azure-configuration
 
 apis:
-- agents
+- responses
 - batches
 - datasetio
 - eval
@@ -11,6 +11,8 @@ apis:
 - safety
 - scoring
 - tool_runtime
+- tool_groups
+- conversations
 - vector_io
 
 benchmarks: []
@@ -51,8 +53,8 @@ providers:
     provider_type: inline::basic
   tool_runtime:
   - config: {} # Enable the RAG tool
-    provider_id: rag-runtime
-    provider_type: inline::rag-runtime
+    provider_id: file-search
+    provider_type: inline::file-search
   - config: {} # Enable MCP (Model Context Protocol) support
     provider_id: model-context-protocol
     provider_type: remote::model-context-protocol
@@ -63,7 +65,7 @@ providers:
         backend: kv_rag
     provider_id: faiss
     provider_type: inline::faiss
-  agents:
+  responses:
   - config:
       persistence:
         agent_state:
@@ -72,8 +74,8 @@ providers:
         responses:
           table_name: agents_responses
           backend: sql_default
-    provider_id: meta-reference
-    provider_type: inline::meta-reference
+    provider_id: builtin
+    provider_type: inline::builtin
   batches:
   - config:
       kvstore:
@@ -99,8 +101,8 @@ providers:
       kvstore:
         namespace: eval_store
         backend: kv_default
-    provider_id: meta-reference
-    provider_type: inline::meta-reference
+    provider_id: builtin
+    provider_type: inline::builtin
 scoring_fns: []
 server:
   port: 8321
@@ -155,8 +157,8 @@ registered_resources:
   scoring_fns: []
   benchmarks: []
   tool_groups:
-  - toolgroup_id: builtin::rag
-    provider_id: rag-runtime
+  - toolgroup_id: builtin::file_search
+    provider_id: file-search
 vector_stores:
   default_provider_id: faiss
   default_embedding_model:

examples/azure-run.yaml

@@ -1,7 +1,7 @@
 version: 2
 
 apis:
-- agents
+- responses
 - batches
 - datasetio
 - eval
@@ -10,6 +10,8 @@ apis:
 - safety
 - scoring
 - tool_runtime
+- tool_groups
+- conversations
 - vector_io
 
 benchmarks: []
@@ -57,8 +59,8 @@ providers:
       openai_api_key: '********'
   tool_runtime:
   - config: {} # Enable the RAG tool
-    provider_id: rag-runtime
-    provider_type: inline::rag-runtime
+    provider_id: file-search
+    provider_type: inline::file-search
   - config: {} # Enable MCP (Model Context Protocol) support
     provider_id: model-context-protocol
     provider_type: remote::model-context-protocol
@@ -69,7 +71,7 @@ providers:
         backend: kv_default
     provider_id: faiss
     provider_type: inline::faiss
-  agents:
+  responses:
   - config:
       persistence:
         agent_state:
@@ -78,8 +80,8 @@ providers:
         responses:
           table_name: agents_responses
           backend: sql_default
-    provider_id: meta-reference
-    provider_type: inline::meta-reference
+    provider_id: builtin
+    provider_type: inline::builtin
   batches:
   - config:
       kvstore:
@@ -105,8 +107,8 @@ providers:
       kvstore:
         namespace: eval_store
         backend: kv_default
-    provider_id: meta-reference
-    provider_type: inline::meta-reference
+    provider_id: builtin
+    provider_type: inline::builtin
 scoring_fns: []
 server:
   port: 8321
@@ -154,8 +156,8 @@ registered_resources:
   scoring_fns: []
   benchmarks: []
   tool_groups:
-  - toolgroup_id: builtin::rag # Register the RAG tool
-    provider_id: rag-runtime
+  - toolgroup_id: builtin::file_search # Register the RAG tool
+    provider_id: file-search
 vector_stores:
   default_provider_id: faiss
   default_embedding_model:

examples/bedrock-run.yaml

@@ -51,8 +51,8 @@ providers:
     provider_type: inline::localfs
   tool_runtime:
   - config: {}
-    provider_id: rag-runtime
-    provider_type: inline::rag-runtime
+    provider_id: file-search
+    provider_type: inline::file-search
   vector_io:
   - config:
       persistence:
@@ -101,11 +101,11 @@ registered_resources:
   shields: []
   vector_stores: []
   tool_groups:
-  - toolgroup_id: builtin::rag
-    provider_id: rag-runtime
+  - toolgroup_id: builtin::file_search
+    provider_id: file-search
 
 # REQUIRED for file_search tool calls to work. Without it, llama-stack's
-# rag-runtime silently fails all file_search operations with no error logged.
+# file-search runtime silently fails all file_search operations with no error logged.
 vector_stores:
   annotation_prompt_params:
     enable_annotations: false

examples/profiles/inline-faiss.yaml

@@ -53,8 +53,8 @@ providers:
     provider_type: inline::llama-guard
   tool_runtime:
   - config: {}
-    provider_id: rag-runtime
-    provider_type: inline::rag-runtime
+    provider_id: file-search
+    provider_type: inline::file-search
   vector_io:
   - config:
       persistence:
@@ -108,11 +108,11 @@ registered_resources:
   shields: []
   vector_stores: []
   tool_groups:
-  - toolgroup_id: builtin::rag
-    provider_id: rag-runtime
+  - toolgroup_id: builtin::file_search
+    provider_id: file-search
 
 # REQUIRED for file_search tool calls to work. Without it, llama-stack's
-# rag-runtime silently fails all file_search operations with no error logged.
+# file-search runtime silently fails all file_search operations with no error logged.
 vector_stores:
   annotation_prompt_params:
     enable_annotations: false

examples/profiles/openai-remote.yaml

@@ -7,7 +7,7 @@
 version: 2
 
 apis:
-- agents
+- responses
 - batches
 - datasetio
 - eval
@@ -16,6 +16,8 @@ apis:
 - safety
 - scoring
 - tool_runtime
+- tool_groups
+- conversations
 - vector_io
 
 benchmarks: []
@@ -59,12 +61,12 @@ providers:
       openai_api_key: '********'
   tool_runtime:
   - config: {} # Enable the RAG tool
-    provider_id: rag-runtime
-    provider_type: inline::rag-runtime
+    provider_id: file-search
+    provider_type: inline::file-search
   - config: {} # Enable the MCP tool
     provider_id: model-context-protocol
     provider_type: remote::model-context-protocol
-  agents:
+  responses:
   - config:
       persistence:
         agent_state:
@@ -73,8 +75,8 @@ providers:
         responses:
           table_name: agents_responses
           backend: sql_default
-    provider_id: meta-reference
-    provider_type: inline::meta-reference
+    provider_id: builtin
+    provider_type: inline::builtin
   batches:
   - config:
       kvstore:
@@ -100,8 +102,8 @@ providers:
       kvstore:
         namespace: eval_store
         backend: kv_default
-    provider_id: meta-reference
-    provider_type: inline::meta-reference
+    provider_id: builtin
+    provider_type: inline::builtin
 scoring_fns: []
 server:
   port: 8321
@@ -142,8 +144,8 @@ registered_resources:
   scoring_fns: []
   benchmarks: []
   tool_groups:
-  - toolgroup_id: builtin::rag # Register the RAG tool
-    provider_id: rag-runtime
+  - toolgroup_id: builtin::file_search # Register the RAG tool
+    provider_id: file-search
 vector_stores:
   annotation_prompt_params: # Override the default Llama Stack annotation that adds <| file-xyz |> to responses
     enable_annotations: true

examples/run.yaml

@@ -2,7 +2,7 @@ version: 2
 image_name: vertexai-configuration
 
 apis:
-- agents
+- responses
 - batches
 - datasetio
 - eval
@@ -11,6 +11,8 @@ apis:
 - safety
 - scoring
 - tool_runtime
+- tool_groups
+- conversations
 - vector_io
 
 benchmarks: []
@@ -51,8 +53,8 @@ providers:
     provider_type: inline::basic
   tool_runtime:
   - config: {} # Enable the RAG tool
-    provider_id: rag-runtime
-    provider_type: inline::rag-runtime
+    provider_id: file-search
+    provider_type: inline::file-search
   - config: {} # Enable MCP (Model Context Protocol) support
     provider_id: model-context-protocol
     provider_type: remote::model-context-protocol
@@ -63,7 +65,7 @@ providers:
         backend: kv_rag
     provider_id: faiss
     provider_type: inline::faiss
-  agents:
+  responses:
   - config:
       persistence:
         agent_state:
@@ -72,8 +74,8 @@ providers:
         responses:
           table_name: agents_responses
           backend: sql_default
-    provider_id: meta-reference
-    provider_type: inline::meta-reference
+    provider_id: builtin
+    provider_type: inline::builtin
   batches:
   - config:
       kvstore:
@@ -99,8 +101,8 @@ providers:
       kvstore:
         namespace: eval_store
         backend: kv_default
-    provider_id: meta-reference
-    provider_type: inline::meta-reference
+    provider_id: builtin
+    provider_type: inline::builtin
 scoring_fns: []
 server:
   port: 8321
@@ -151,8 +153,8 @@ registered_resources:
   scoring_fns: []
   benchmarks: []
   tool_groups:
-  - toolgroup_id: builtin::rag
-    provider_id: rag-runtime
+  - toolgroup_id: builtin::file_search
+    provider_id: file-search
 vector_stores:
   default_provider_id: faiss
   default_embedding_model:

examples/vertexai-run.yaml

@@ -52,8 +52,8 @@ providers:
     provider_type: inline::basic
   tool_runtime:
   - config: {}
-    provider_id: rag-runtime
-    provider_type: inline::rag-runtime
+    provider_id: file-search
+    provider_type: inline::file-search
   vector_io:
   - config:
       persistence:
@@ -139,8 +139,8 @@ registered_resources:
   scoring_fns: []
   benchmarks: []
   tool_groups:
-  - toolgroup_id: builtin::rag
-    provider_id: rag-runtime
+  - toolgroup_id: builtin::file_search
+    provider_id: file-search
 vector_stores:
   default_provider_id: faiss
   default_embedding_model:

examples/vllm-rhaiis.yaml

@@ -60,8 +60,8 @@ providers:
       openai_api_key: '********'
   tool_runtime:
   - config: {} # Enable the RAG tool
-    provider_id: rag-runtime
-    provider_type: inline::rag-runtime
+    provider_id: file-search
+    provider_type: inline::file-search
   - config: {} # Enable MCP (Model Context Protocol) support
     provider_id: model-context-protocol
     provider_type: remote::model-context-protocol
@@ -151,8 +151,8 @@ registered_resources:
   scoring_fns: []
   benchmarks: []
   tool_groups:
-  - toolgroup_id: builtin::rag # Register the RAG tool
-    provider_id: rag-runtime
+  - toolgroup_id: builtin::file_search # Register the RAG tool
+    provider_id: file-search
 vector_stores:
   default_provider_id: faiss
   default_embedding_model: # Define the default embedding model for RAG