Merge branch 'master' into ignore_file

README.md

@@ -35,7 +35,8 @@ The mono-repo for the Couchbase Agent Catalog project.
    git clone [email protected]:couchbaselabs/agent-catalog.git
    ```
 
-3. You are now ready to install the Agent Catalog package!
+3. You are now ready to install the Agent Catalog package! We recommend using Anaconda to create a virtual environment
+   for your project to ensure no global dependencies interfere with the project.
    Using your project's Python environment, execute the following command to install a local package with `pip`:
 
    ```bash
@@ -45,6 +46,11 @@ The mono-repo for the Couchbase Agent Catalog project.
    # Install the agentc package.
    pip install libs/agentc
    ```
+   If you are interested in developing with langchain, also install `agentc_langchain` by running the following:
+
+   ```bash
+   pip install libs/agentc_langchain
+   ```
 
    If you are interested in building a ``.whl`` file (for later use in ``.whl``-based installs), use :command:`poetry`
    directly:
@@ -107,12 +113,12 @@ Commands:
   execute  Search and execute a specific tool.
   find     Find items from the catalog based on a natural language QUERY string or by name.
   index    Walk the source directory trees (SOURCE_DIRS) to index source files into the local catalog.
-  ls       List all tools or prompts in the catalog.
-  publish  Upload the local catalog to a Couchbase instance.
+  ls       List all indexed tools and/or prompts in the catalog.
+  publish  Upload the local catalog and/or logs to a Couchbase instance.
   status   Show the status of the local catalog.
   version  Show the current version of agentc.
 
-  See: https://docs.couchbase.com for more information.
+  See: https://docs.couchbase.com or https://couchbaselabs.github.io/agent-catalog/index.html# for more information.
 ```
 
 If you see the output above, you are all set! Head on over to our [docs](https://couchbaselabs.github.io/agent-catalog/) or our [templates](templates) to start

libs/agentc_cli/agentc_cli/main.py

@@ -74,7 +74,7 @@ def resolve_command(self, ctx, args):
 
 @click.group(
     cls=AliasedGroup,
-    epilog="See: https://docs.couchbase.com for more information.",
+    epilog="See: https://docs.couchbase.com or https://couchbaselabs.github.io/agent-catalog/index.html# for more information.",
     context_settings=dict(max_content_width=800),
 )
 @click.option(

libs/agentc_core/agentc_core/catalog/catalog/db.py

@@ -63,7 +63,9 @@ def find(
         """Returns the catalog items that best match a query."""
 
         # Catalog item has to be queried directly
-        if name is not None and snapshot is not None:
+        if name is not None:
+            if snapshot == LATEST_SNAPSHOT_VERSION:
+                snapshot = self.version.identifier
             # TODO (GLENN): Need to add some validation around bucket (to prevent injection)
             # TODO (GLENN): Need to add some validation around name (to prevent injection)
             item_query = f"""
@@ -77,18 +79,6 @@ def find(
                 return []
             query_embeddings = None
 
-        elif name is not None:
-            item_query = f"""
-                FROM `{self.bucket}`.`{DEFAULT_CATALOG_SCOPE}`.`{self.kind}_catalog` AS a
-                WHERE a.name = $name
-                SELECT a.*;
-            """
-            res, err = execute_query_with_parameters(self.cluster, item_query, {"name": name})
-            if err is not None:
-                logger.error(err)
-                return []
-            query_embeddings = None
-
         else:
             # Generate embeddings for user query
             query_embeddings = self.embedding_model.encode(query)
@@ -202,6 +192,9 @@ def find(
             descriptors.append(descriptor)
 
         # We compute the true cosine distance here (Couchbase uses a different score :-)).
+        if name is not None:
+            return [SearchResult(entry=descriptors[0], delta=1)]
+
         deltas = self.get_deltas(query_embeddings, [t.embedding for t in descriptors])
         results = [SearchResult(entry=descriptors[i], delta=deltas[i]) for i in range(len(deltas))]
         return sorted(results, key=lambda t: t.delta, reverse=True)

libs/agentc_core/tests/README.md

@@ -1,5 +1,5 @@
 # Agent-Catalog Core Tests
 
 1. Set your current working directory to be the parent folder (`libs/agentc_core`).
-2. Activate your Poetry environment with `poetry shell`.
+2. Activate your Poetry environment.
 3. Run `pytest tests` and wait for the results! To run just the smoke tests, use `pytest tests -m smoke`.

templates/agents/with_controlflow/README.md

@@ -27,7 +27,6 @@ This directory contains a starter project for building agents with Couchbase, Co
 
    ```bash
    poetry install --with analysis
-   poetry shell
    ```
 
 5. Run `agentc` to make sure this project has installed correctly (note that your first run will take a couple of
@@ -57,7 +56,7 @@ This directory contains a starter project for building agents with Couchbase, Co
      status   Show the status of the local catalog.
      version  Show the current version of agentc.
 
-     See: https://docs.couchbase.com for more information.
+     See: https://docs.couchbase.com or https://couchbaselabs.github.io/agent-catalog/index.html# for more information.
    ```
 
 6. Make sure your Git repo is clean, and run `agentc index` to index your tools and prompts.