agent-cli 0.72.7__py3-none-any.whl → 0.73.0__py3-none-any.whl

agent_cli/agents/memory/proxy.py

@@ -23,6 +23,7 @@ def proxy(
         rich_help_panel="Memory Configuration",
     ),
     openai_base_url: str | None = opts.OPENAI_BASE_URL,
+    embedding_base_url: str | None = opts.EMBEDDING_BASE_URL,
     embedding_model: str = opts.EMBEDDING_MODEL,
     openai_api_key: str | None = opts.OPENAI_API_KEY,
     default_top_k: int = typer.Option(
@@ -135,12 +136,15 @@ def proxy(
     entries_dir, _ = ensure_store_dirs(memory_path)
     if openai_base_url is None:
         openai_base_url = constants.DEFAULT_OPENAI_BASE_URL
+    effective_embedding_url = embedding_base_url or openai_base_url
 
     console.print(f"[bold green]Starting Memory Proxy on {host}:{port}[/bold green]")
     console.print(f"  💾 Memory store: [blue]{memory_path}[/blue]")
     console.print(f"  📁 Entries: [blue]{entries_dir}[/blue]")
     console.print(f"  🤖 Backend: [blue]{openai_base_url}[/blue]")
-    console.print(f"  🧠 Embeddings: Using [blue]{embedding_model}[/blue]")
+    console.print(
+        f"  🧠 Embeddings: [blue]{embedding_model}[/blue] via [blue]{effective_embedding_url}[/blue]",
+    )
     console.print(f"  🔍 Memory top_k: [blue]{default_top_k}[/blue] entries per query")
     console.print(f"  🧹 Max entries per conversation: [blue]{max_entries}[/blue]")
     console.print(
@@ -154,6 +158,7 @@ def proxy(
     fastapi_app = create_app(
         memory_path,
         openai_base_url,
+        embedding_base_url=effective_embedding_url,
         embedding_model=embedding_model,
         embedding_api_key=openai_api_key,
         chat_api_key=openai_api_key,

agent_cli/agents/rag_proxy.py

@@ -32,6 +32,7 @@ def rag_proxy(
         rich_help_panel="RAG Configuration",
     ),
     openai_base_url: str | None = opts.OPENAI_BASE_URL,
+    embedding_base_url: str | None = opts.EMBEDDING_BASE_URL,
     embedding_model: str = opts.EMBEDDING_MODEL,
     openai_api_key: str | None = opts.OPENAI_API_KEY,
     limit: int = typer.Option(
@@ -138,22 +139,26 @@ def rag_proxy(
 
     if openai_base_url is None:
         openai_base_url = constants.DEFAULT_OPENAI_BASE_URL
+    effective_embedding_url = embedding_base_url or openai_base_url
 
     console.print(f"[bold green]Starting RAG Proxy on {host}:{port}[/bold green]")
     console.print(f"  📂 Docs: [blue]{docs_folder}[/blue]")
     console.print(f"  💾 DB: [blue]{chroma_path}[/blue]")
     console.print(f"  🤖 Backend: [blue]{openai_base_url}[/blue]")
-    console.print(f"  🧠 Embeddings: Using [blue]{embedding_model}[/blue]")
+    console.print(
+        f"  🧠 Embeddings: [blue]{embedding_model}[/blue] via [blue]{effective_embedding_url}[/blue]",
+    )
     console.print(f"  🔍 Limit: [blue]{limit}[/blue] chunks per query")
 
     fastapi_app = create_app(
         docs_folder,
         chroma_path,
         openai_base_url,
-        embedding_model,
-        openai_api_key,
-        openai_api_key,
-        limit,
+        embedding_base_url=effective_embedding_url,
+        embedding_model=embedding_model,
+        embedding_api_key=openai_api_key,
+        chat_api_key=openai_api_key,
+        limit=limit,
         enable_rag_tools=enable_rag_tools,
     )
 

agent_cli/memory/api.py

@@ -22,6 +22,7 @@ LOGGER = logging.getLogger(__name__)
 def create_app(
     memory_path: Path,
     openai_base_url: str,
+    embedding_base_url: str | None = None,
     embedding_model: str = DEFAULT_OPENAI_EMBEDDING_MODEL,
     embedding_api_key: str | None = None,
     chat_api_key: str | None = None,
@@ -39,6 +40,7 @@ def create_app(
     client = MemoryClient(
         memory_path=memory_path,
         openai_base_url=openai_base_url,
+        embedding_base_url=embedding_base_url,
         embedding_model=embedding_model,
         embedding_api_key=embedding_api_key,
         chat_api_key=chat_api_key,

agent_cli/memory/client.py

@@ -41,6 +41,7 @@ class MemoryClient:
         self,
         memory_path: Path,
         openai_base_url: str,
+        embedding_base_url: str | None = None,
         embedding_model: str = DEFAULT_OPENAI_EMBEDDING_MODEL,
         embedding_api_key: str | None = None,
         chat_api_key: str | None = None,
@@ -74,7 +75,7 @@ class MemoryClient:
         self.collection: Collection = init_memory_collection(
             self.memory_path,
             embedding_model=embedding_model,
-            openai_base_url=self.openai_base_url,
+            openai_base_url=embedding_base_url or self.openai_base_url,
             openai_api_key=embedding_api_key,
         )
 

agent_cli/opts.py

@@ -109,6 +109,14 @@ EMBEDDING_MODEL: str = typer.Option(
     help="Embedding model to use for vectorization.",
     rich_help_panel="LLM Configuration",
 )
+EMBEDDING_BASE_URL: str | None = typer.Option(
+    None,
+    "--embedding-base-url",
+    envvar="EMBEDDING_BASE_URL",
+    help="Base URL for embedding API. Falls back to `--openai-base-url` if not set. "
+    "Useful when using different providers for chat vs embeddings.",
+    rich_help_panel="LLM Configuration",
+)
 
 # --- ASR (Audio) Configuration ---
 # General ASR

agent_cli/rag/api.py

@@ -32,6 +32,7 @@ def create_app(
     docs_folder: Path,
     chroma_path: Path,
     openai_base_url: str,
+    embedding_base_url: str | None = None,
     embedding_model: str = DEFAULT_OPENAI_EMBEDDING_MODEL,
     embedding_api_key: str | None = None,
     chat_api_key: str | None = None,
@@ -47,7 +48,7 @@ def create_app(
         chroma_path,
         name="docs",
         embedding_model=embedding_model,
-        openai_base_url=openai_base_url,
+        openai_base_url=embedding_base_url or openai_base_url,
         openai_api_key=embedding_api_key,
     )
 

{agent_cli-0.72.7.dist-info → agent_cli-0.73.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agent-cli
-Version: 0.72.7
+Version: 0.73.0
 Summary: A suite of AI-powered command-line tools for text correction, audio transcription, and voice assistance.
 Project-URL: Homepage, https://github.com/basnijholt/agent-cli
 Author-email: Bas Nijholt <bas@nijho.lt>
@@ -2066,8 +2066,12 @@ uv tool install "agent-cli[vad]" -p 3.13
 │                                [env var: OPENAI_API_KEY]                               │
 ╰────────────────────────────────────────────────────────────────────────────────────────╯
 ╭─ LLM Configuration ────────────────────────────────────────────────────────────────────╮
-│ --embedding-model        TEXT  Embedding model to use for vectorization.               │
-│                                [default: text-embedding-3-small]                       │
+│ --embedding-base-url        TEXT  Base URL for embedding API. Falls back to            │
+│                                   --openai-base-url if not set. Useful when using      │
+│                                   different providers for chat vs embeddings.          │
+│                                   [env var: EMBEDDING_BASE_URL]                        │
+│ --embedding-model           TEXT  Embedding model to use for vectorization.            │
+│                                   [default: text-embedding-3-small]                    │
 ╰────────────────────────────────────────────────────────────────────────────────────────╯
 ╭─ Server Configuration ─────────────────────────────────────────────────────────────────╮
 │ --host        TEXT     Host/IP to bind API servers to.                                 │
@@ -2229,8 +2233,12 @@ The `memory proxy` command is the core feature—a middleware server that gives
 │                                [env var: OPENAI_API_KEY]                               │
 ╰────────────────────────────────────────────────────────────────────────────────────────╯
 ╭─ LLM Configuration ────────────────────────────────────────────────────────────────────╮
-│ --embedding-model        TEXT  Embedding model to use for vectorization.               │
-│                                [default: text-embedding-3-small]                       │
+│ --embedding-base-url        TEXT  Base URL for embedding API. Falls back to            │
+│                                   --openai-base-url if not set. Useful when using      │
+│                                   different providers for chat vs embeddings.          │
+│                                   [env var: EMBEDDING_BASE_URL]                        │
+│ --embedding-model           TEXT  Embedding model to use for vectorization.            │
+│                                   [default: text-embedding-3-small]                    │
 ╰────────────────────────────────────────────────────────────────────────────────────────╯
 ╭─ Server Configuration ─────────────────────────────────────────────────────────────────╮
 │ --host        TEXT     Host/IP to bind API servers to.                                 │

{agent_cli-0.72.7.dist-info → agent_cli-0.73.0.dist-info}/RECORD

@@ -9,7 +9,7 @@ agent_cli/config_cmd.py,sha256=PkFY-U09LRIFYrHL_kG4_Ge6DjCWFe3GkO_uiIBMTgI,10359
 agent_cli/constants.py,sha256=-Q17N6qKIGqPDsu3FxpIKP33G0Cs0RUJlHwYNHxVxek,843
 agent_cli/docs_gen.py,sha256=ZX2GYHNumpLhdAEc_4Hy6xeAahAzsEVVnsXUojMYVNY,13885
 agent_cli/example-config.toml,sha256=xd9BXeOqdYx4xFJt58VBs2I49ESy6dF4-mWF_g8sM9o,7552
-agent_cli/opts.py,sha256=BzPIZrGkaoHXw83imRbxKNZJVwJzbpVGcmrpQavTesk,13089
+agent_cli/opts.py,sha256=XgVFv-ip5lkFJNyBGHewCBQc4YaLJUSijIsP1qiqcts,13405
 agent_cli/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 agent_cli/_requirements/.gitkeep,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 agent_cli/_requirements/audio.txt,sha256=KAOtFaFClkj2EAesWI61LZ4N-T1WQvHtvf1sNtDxVgQ,1522
@@ -29,14 +29,14 @@ agent_cli/agents/_voice_agent_common.py,sha256=PUAztW84Xf9U7d0C_K5cL7I8OANIE1H6M
 agent_cli/agents/assistant.py,sha256=oUu6Z3mRQnO_PKzby71iB282UN-yRmKGiPa-OzkHmRg,13794
 agent_cli/agents/autocorrect.py,sha256=nJnNC9uKBZcbQ4roiC-BiJKTomiSdfe8mNa8p3yACps,9715
 agent_cli/agents/chat.py,sha256=DhG7qAYqlwj_UDaMREucFhvUqNuWNLFGpR_NRp7nuek,18703
-agent_cli/agents/rag_proxy.py,sha256=0-CEgthAqQ3bQ4NISiS0fPjjkAW44KNylChenXSjW5A,6273
+agent_cli/agents/rag_proxy.py,sha256=0EeCbMqnFAGuCKe1wtjaYJsfbMlUfFxN0-uekYzLWpg,6560
 agent_cli/agents/speak.py,sha256=XcBK_y1LMdNPdgfs9V57FNSZMgtjWFogIPsUx0g4SX0,7640
 agent_cli/agents/transcribe.py,sha256=4KrMsNty4O2sRtmhqsHzH5AYXoMeTrbFq0mPrV_qGqE,25322
 agent_cli/agents/transcribe_daemon.py,sha256=kJhmp7143fj3mhwz-l-a-LKvxZkhcGMHHcn8Cvq-Ec4,18926
 agent_cli/agents/voice_edit.py,sha256=OiYgBJ9GhtbeE7uLfaGHLWHXe7TKNd9Oo6kfvsROJ3M,11405
 agent_cli/agents/memory/__init__.py,sha256=3IFV1fSD0u38E54f_0oPZoPbM0LXaH-A3bDSnBL9FJs,1470
 agent_cli/agents/memory/add.py,sha256=i_F78rnE8ftJ1DY1yX4BpClV8H9Rz9-aFh6lzI6a_zE,6349
-agent_cli/agents/memory/proxy.py,sha256=NlEAg6sgr0SBRObfOmqV2PSONA6rcOTBua2rTSPny6U,6975
+agent_cli/agents/memory/proxy.py,sha256=rSesjHTx366Q9fEXB-DanRpFOJnua3tkQwXkJeMmRgs,7209
 agent_cli/core/__init__.py,sha256=c_knH7u9QgjsfMIil9NP4bVizHawLUMYoQWU4H9vMlQ,46
 agent_cli/core/audio.py,sha256=43FpYe2Wu_BYK9xJ_55V4xHjHJeFwQ5aM-CQzlTryt8,15168
 agent_cli/core/audio_format.py,sha256=zk3qlYMAlKYPz1enrjihQQspl_C218v1Rbcm7Uktlew,8773
@@ -107,8 +107,8 @@ agent_cli/memory/_retrieval.py,sha256=K_2TUcgzfntBARPyf0K6VR3NIgHHJrqGFMP_53Nae_
 agent_cli/memory/_store.py,sha256=m9mD1GxjdTXpnyL-X-MIU4cj28unqxJ_azV3kwM8blM,5086
 agent_cli/memory/_streaming.py,sha256=P1JnkDNTJJj-lXawmXhBZnIia3ZZmKo-N6mUsLYFVgs,1400
 agent_cli/memory/_tasks.py,sha256=XgEkN_3NCVQDWafZ_rqazpAE68yQ87x-amnQKMkfPXg,1469
-agent_cli/memory/api.py,sha256=riaGTOIw7g3KuDNcPtDy9wO44-D1wDtVadRj3RTT10k,3595
-agent_cli/memory/client.py,sha256=XomHhP-hPSoosORkBKSY1dW3gjheFEVqAac0b-tAULo,9994
+agent_cli/memory/api.py,sha256=h7sxbNsq72ewOTnv4XcHaMBGY5NgEMAB5e_OmtMzxy0,3685
+agent_cli/memory/client.py,sha256=szkI8G8CHfHOgH9LwDEQQmZZNsUi_a3DdHX23h_yn2U,10063
 agent_cli/memory/engine.py,sha256=rABVC86b5wU1QxY3BM43RhvfDOxoRT7Ddm98BN_qCL4,11656
 agent_cli/memory/entities.py,sha256=_8wyJz--tNa66CEtSpl2TUN_zeHQvMzm42htnDaOr6g,1219
 agent_cli/memory/models.py,sha256=KK0wToEf-tXssYVL0hYaJlcADlJ3G2lcSXwo1UmA0VU,2352
@@ -119,7 +119,7 @@ agent_cli/rag/_prompt.py,sha256=d8_jOhZGafMmjO7BlCl4H125bj4m-dNFWDOLz5_OPrw,954
 agent_cli/rag/_retriever.py,sha256=bzMzZR43P5cROgnWwOh_BrMFsMP5tDm21ToFVZwb0gk,4505
 agent_cli/rag/_store.py,sha256=HksCLnbHp19dnY5ZWglm86azBjjuiWqvZvRPG-oJ8SY,1381
 agent_cli/rag/_utils.py,sha256=OKZvn8UFb3TsB4b0eIWU6Md5xDiaG-g659zMjVUu8oI,5923
-agent_cli/rag/api.py,sha256=Q_aaRdNxh9avNGvXWEJ5qH-vMTagN7m-iJdTZSeSink,5564
+agent_cli/rag/api.py,sha256=s0jsCrACkuWfTiGzznlacxyTGGu9Fa6LVmrIf23JBbQ,5629
 agent_cli/rag/client.py,sha256=mFiZ4yjI75Vsehie6alsV1My50uIsp-G07Qz6SaNrZw,8913
 agent_cli/rag/engine.py,sha256=XySDer0fNTsEUjbUby5yf7JqB7uCE7tw2A6tYJixHnI,9800
 agent_cli/rag/models.py,sha256=uECWoeBChlkAK7uTM-pUnPGaaMO4EYJ3pJcAf8uh1vI,1043
@@ -189,8 +189,8 @@ agent_cli/services/asr.py,sha256=aRaCLVCygsJ15qyQEPECOZsdSrnlLPbyY4RwAqY0qIw,172
 agent_cli/services/llm.py,sha256=i01utl1eYWlM13gvW2eR6ErL_ndH-g0d-BSleZra_7k,7229
 agent_cli/services/tts.py,sha256=NX5Qnq7ddLI3mwm3nzhbR3zB1Os4Ip4sSVSjDZDTBcI,14855
 agent_cli/services/wake_word.py,sha256=JFJ1SA22H4yko9DXiQ1t7fcoxeALLAe3iBrLs0z8rX4,4664
-agent_cli-0.72.7.dist-info/METADATA,sha256=F1HKnb3c-lJozyL0wkUqxhOFQvNlCTEwO3ZtM3IcPLo,179913
-agent_cli-0.72.7.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
-agent_cli-0.72.7.dist-info/entry_points.txt,sha256=FUv-fB2atLsPUk_RT4zqnZl1coz4_XHFwRALOKOF38s,97
-agent_cli-0.72.7.dist-info/licenses/LICENSE,sha256=majJU6S9kC8R8bW39NVBHyv32Dq50FL6TDxECG2WVts,1068
-agent_cli-0.72.7.dist-info/RECORD,,
+agent_cli-0.73.0.dist-info/METADATA,sha256=ybJ--hEZOx6RMgWCcWysDtu-eiMsRMq5H74VsaTYrMQ,180673
+agent_cli-0.73.0.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
+agent_cli-0.73.0.dist-info/entry_points.txt,sha256=FUv-fB2atLsPUk_RT4zqnZl1coz4_XHFwRALOKOF38s,97
+agent_cli-0.73.0.dist-info/licenses/LICENSE,sha256=majJU6S9kC8R8bW39NVBHyv32Dq50FL6TDxECG2WVts,1068
+agent_cli-0.73.0.dist-info/RECORD,,