## Description
This sets the default annotation (readOnlyHint, destructiveHint, etc.)
in the looker tool code.
If desired, a custom tool can override the annotation in yaml like
```yaml
tool_name:
kind: tool-kind
source: tool-source
annotations:
readOnlyHint: false
destructiveHint: true
description: |
Tool Description
```
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [x] Appropriate docs were updated (if necessary)
- [x] Make sure to add `!` if this involve a breaking change
Co-authored-by: Averi Kitsch <akitsch@google.com>
## Description
This Pull Request addresses an issue where the `postgres-list-tables`
tool was including the `google_ml` schema in the results when listing
tables from an AlloyDB PostgreSQL instance.
The primary change involves updating the SQL query within
[https://github.com/googleapis/genai-toolbox/blob/main/internal/tools/postgres/postgreslisttables/postgreslisttables.go](url)
to explicitly exclude the `google_ml` schema, alongside other system
schemas.
The `WHERE` clause in the `listTablesStatement` has been modified from:
```sql
-- Original WHERE clause snippet
AND ns.nspname NOT IN ('pg_catalog', 'information_schema', 'pg_toast')
AND ns.nspname NOT LIKE 'pg_temp_%' AND ns.nspname NOT LIKE 'pg_toast_temp_%'
```
To:
```sql
-- Updated WHERE clause snippet
AND ns.nspname NOT IN ('pg_catalog', 'information_schema', 'pg_toast','google_ml')
AND ns.nspname NOT LIKE 'pg_temp_%' AND ns.nspname NOT LIKE 'pg_toast_temp_%'
```
This ensures that tables residing in the google_ml schema are no longer
returned by the tool, improving the clarity and relevance of the table
listings for users.
## Testing
The changes were validated end-to-end by:
* Building a custom `toolbox` binary from the modified source.
* Configuring `~/.gemini/settings.json` to use this binary as an
`alloydb-test` MCP server.
* Launching `gemini` CLI and confirming the `alloydb-test` server was `🟢
Ready` via `/mcp`.
* Invoking the `postgres-list-tables` tool with `list tables`.
* Verifying that `google_ml` schema tables were successfully excluded
from the results.
Testing screenshot :
<img width="2809" height="1779" alt="Screenshot 2025-11-25 3 47 43 PM"
src="https://github.com/user-attachments/assets/126cf5be-30d7-4ec1-9d31-652b5219c0ce"
/>
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [ ] Code coverage does not decrease (if any source code was changed)
- [ ] Appropriate docs were updated (if necessary)
- [ ] Make sure to add `!` if this involve a breaking change
🛠️Fixes#2009
Co-authored-by: prernakakkar-google <158031829+prernakakkar-google@users.noreply.github.com>
## Description
---
This pull request adds a new tool, cloud-sql-clone-instance, which
enables cloning a Cloud SQL instance from the toolbox using the Cloud
SQL Admin API. The tool supports both standard cloning and point-in-time
recovery (PITR). It also supports specifying preferred zones for cloned
instances via the preferredZone and preferredSecondaryZone fields.
Key Features:
Instance Cloning: The tool allows you to clone a Cloud SQL instance by
specifying the source and destination instance names.
Point-in-Time Recovery (PITR): By providing a pointInTime timestamp, you
can create a clone of an instance as it existed at a specific moment.
High Availability Configuration: The preferredZone and
preferredSecondaryZone parameters allow you to configure the cloned
instance for high availability.
Tested:
<img width="1182" height="446" alt="Screenshot 2025-11-11 at 12 21
47 PM"
src="https://github.com/user-attachments/assets/7f39a5a3-3967-43d0-8041-f1d47b4fbcd9"
/>
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [x] Appropriate docs were updated (if necessary)
- [ ] Make sure to add `!` if this involve a breaking change
🛠️Fixes#1915
Co-authored-by: prernakakkar-google <158031829+prernakakkar-google@users.noreply.github.com>
Co-authored-by: Averi Kitsch <akitsch@google.com>
## Description
This pull request resolves issue
[#36](https://github.com/gemini-cli-extensions/sql-server/issues/36) by
introducing default connection parameters for the MSSQL prebuilt tool.
If a user does not specify a host or port, the tool will now default to
localhost:1433
**Changes Implemented**
The modification is in internal/prebuiltconfigs/tools/mssql.yaml, where
the host and port fields now include default values:
host: ${MSSQL_HOST:localhost}
port: ${MSSQL_PORT:1433}
This configuration allows the tool to function using the defaults, but
users can still set the MSSQL_HOST or MSSQL_PORT environment variables
to override them.
**Validation Process**
Validated changes by running the toolbox against a Microsoft SQL Server
instance hosted in a Docker container.
- **Database Setup:** A testdb database containing a products table was
initialized.
- **CLI Configuration:** The ~/.gemini/settings.json file was updated to
point to my local toolbox build.
- **Tool Launch:** The UI was started using go run . --prebuilt mssql
--ui
- **Testing:** Confirmed the connection logic by testing two scenarios.
One with the environment variables set and one without (to confirm the
default logic).
**1.** Default Value Connection: For this test, the MSSQL_HOST and
MSSQL_PORT environment variables were not defined. The application
correctly utilized the new default values (localhost:1433) to connect to
the testdb.
<img width="2175" height="1144" alt="Screenshot 2025-11-12 6 33 05 PM"
src="https://github.com/user-attachments/assets/bacfc9bf-8b35-42e1-ad53-4af3aef27125"
/>
**2.** Explicit Variable Connection: For the second test, MSSQL_HOST and
MSSQL_PORT were set to specific values. The application correctly
prioritized these variables over the new defaults and connected
successfully.
<img width="2175" height="1144" alt="Screenshot 2025-11-12 6 37 02 PM"
src="https://github.com/user-attachments/assets/61254849-211d-41f4-8c7d-ff92cf64a51c"
/>
Result: Both connection methods were verified by executing the
list_tables prebuilt tool via the UI. In each scenario, the tool
successfully retrieved the UserDetails table, confirming the changes.
<img width="2250" height="1144" alt="Screenshot 2025-11-12 6 38 12 PM"
src="https://github.com/user-attachments/assets/87085777-897e-4a74-9e3f-f36cc8a33305"
/>
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [ ] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [ ] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [ ] Ensure the tests and linter pass
- [ ] Code coverage does not decrease (if any source code was changed)
- [ ] Appropriate docs were updated (if necessary)
- [ ] Make sure to add `!` if this involve a breaking change
🛠️ Fixes
[#36](https://github.com/gemini-cli-extensions/sql-server/issues/36)
## Description
Tool `invoke()` and `RequiresClientAuthorization()` takes a new input
argument -- Resource Manager. Resource manager will be used to retrieve
Source in the next step.
In order to achieve the goal, this PR implements the follows:
* move resource manager from the server package to a new package to
prevent import cycles (between server and mcp)
* added a new interface in `tools.go` to prevent import cycle (between
resources and tools package)
* add new input argument in all tools
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [x] Appropriate docs were updated (if necessary)
- [x] Make sure to add `!` if this involve a breaking change
## Description
Add the import statement for firebirdsql. Our integration test runs on
linux and did not catch this error.
Seems like windows will look for the import statement -- `unable to
create pool: unable to create connection pool: sql: unknown driver
\"firebirdsql\" (forgotten import?)"`
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [x] Appropriate docs were updated (if necessary)
- [x] Make sure to add `!` if this involve a breaking change
🛠️Fixes#2014
## Description
Update error log to more details when initialization fails.
* When a toolset name is not valid, show the toolset name (similar with
promptset).
* When tool does not exist, show the tool name that is missing (similar
with prompt).
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [x] Appropriate docs were updated (if necessary)
- [x] Make sure to add `!` if this involve a breaking change
🛠️Fixes#1579
Support `allowed-origins` flag to allow secure deployment of Toolbox.
Current Toolbox is **insecure by default**, which allows all origin
(`*`). This PR also updated docs to notify user of the new
`allowed-origins` flag in the Cloud Run, kubernetes, and docker
deployment docs.
This PR was tested manually by mocking a browser access:
1. Created a HTML file with Javascript fetch named
`malicious-client.html`:
```
<!DOCTYPE html>
<html>
<head>
<title>Malicious CORS Test</title>
</head>
<body>
<h1>Attempting to access API at http://127.0.0.1:5000/mcp</h1>
<p>Check the **Chrome Developer Console** (F12 -> Console tab) for the result.</p>
<script>
fetch('http://127.0.0.1:5000/mcp', {
method: 'POST',
headers: {
'Content-Type': 'application/json',
// The browser automatically adds the 'Origin' header based on where this HTML is served from (http://localhost:8000)
},
body: JSON.stringify({
"jsonrpc": "2.0",
"id": 1,
"method": "tools/list"
})
})
.then(response => {
console.log('Success (but check console for CORS enforcement details):', response);
return response.json();
})
.then(data => console.log('Data received (only if CORS passes):', data))
.catch(error => console.error('Fetch Error:', error));
</script>
</body>
</html>
```
2. Run `python3 -m http.server 8000`
3. Open `http://localhost:8000/malicious-client.html` in browser.
4. Tried without `--allowed-origins` flag -- success.
Tried with `--allowed-origins=http://localhost:8000` -- success.
Tried with `--allowed-origins=http://foo.com` -- unsuccessful.
---------
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Averi Kitsch <akitsch@google.com>
## Description
The MCP spec supports tool annotations like the below structure in the
2025-06-18 version of the spec.
https://modelcontextprotocol.io/specification/2025-06-18/schema#toolannotations
```
{
destructiveHint?: boolean;
idempotentHint?: boolean;
openWorldHint?: boolean;
readOnlyHint?: boolean;
}
```
Added a ToolAnnotations structure, an Annotations member to the
McpManifest structure, and a nil initializer for the Annotations member
to all calls to GetMcpManifest.
The ToolAnnotations structure and the member annotations are all defined
as pointers so that they are omited when not set. There are times when
the zero value is meaningful so this was the only way to make sure that
we distinguish between not setting the annotation and setting it with a
zero value.
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [x] Appropriate docs were updated (if necessary)
- [x] Make sure to add `!` if this involve a breaking change
🛠️Fixes#927
---------
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
## Description
Spanner List Graphs tool, similar to list tables it can be used to get
all/specific graph details
## PR Checklist
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [x] Appropriate docs were updated (if necessary)
- [x] Make sure to add `!` if this involve a breaking change
🛠️Fixes#1916
---------
Co-authored-by: Averi Kitsch <akitsch@google.com>
## Description
Implements the 'postgres-upgrade-precheck' tool to allow users to
validate instance readiness for major version upgrades for CloudSQL
PostgreSQL.
This includes the tool implementation, unit tests for YAML parsing,
integration tests for tool invocation, and documentation. The tool is
also added to the CloudSQL PostgreSQL prebuilt set.
TEST output:
<img width="3406" height="1646" alt="image"
src="https://github.com/user-attachments/assets/6abaa535-285d-4645-9dd3-7ebcd447d448"
/>
<img width="3532" height="1490" alt="image"
src="https://github.com/user-attachments/assets/4d512af1-51fd-4187-b80f-be13198aba68"
/>
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [x] Appropriate docs were updated (if necessary)
- [x] Make sure to add `!` if this involve a breaking change
🛠️Fixes#1721
---------
Co-authored-by: Averi Kitsch <akitsch@google.com>
## Description
> Should include a concise description of the changes (bug or feature),
it's
> impact, along with a summary of the solution
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [ ] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [ ] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [ ] Ensure the tests and linter pass
- [ ] Code coverage does not decrease (if any source code was changed)
- [ ] Appropriate docs were updated (if necessary)
- [ ] Make sure to add `!` if this involve a breaking change
🛠️ Fixes #<issue_number_goes_here>
---------
## Description
This commit allows a tool to pull an alternate authorization
token from the header of the http request.
This is initially being built for the Looker integration. Looker
uses its own OAuth token. When deploying MCP Toolbox to Cloud
Run, the default token in the "Authorization" header is for
authentication with Cloud Run. An alternate token can be put into
another header by a client such as ADK or any other client that
can programatically set http headers. This token will be used
to authenticate with Looker.
If needed, other sources can use this by setting the header name
in the source config, passing it into the tool config, and returning
the header name in the Tool GetAuthTokenHeaderName() function.
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [x] Appropriate docs were updated (if necessary)
- [x] Make sure to add `!` if this involve a breaking change
🛠️Fixes#1540
## Description
Previous merge conflict caused parameter duplications
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [x] Appropriate docs were updated (if necessary)
- [x] Make sure to add `!` if this involve a breaking change
🛠️ Fixes https://github.com/gemini-cli-extensions/alloydb/issues/97
To keep a persistent backend storage for configuration, we will have to
keep a single source of truth. This involves supporting bi-directional
conversion for Prompt and Promptsets.
This PR make the following changes:
* Separate Prompt from PromptConfig
* Embed Config in Resources (for both prompt and promptset)
* Add `ToConfig()` to extract Config from Resources (for both prompt and
promptset)
## Description
Return error if there's error during query execution. AlloyDB AI NL tool
was failing query silently. When this happened, the user was just
receiving `nil` response and not informed of what the error was.
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [x] Appropriate docs were updated (if necessary)
- [x] Make sure to add `!` if this involve a breaking change
🛠️Fixes#1949
## Description
Unmarshal object_details json string into map to make response have
nested json in spannerlisttables instead of nested JSON string
## PR Checklist
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [x] Appropriate docs were updated (if necessary)
- [x] Make sure to add `!` if this involve a breaking change
🛠️Fixes#1838
---------
Co-authored-by: Averi Kitsch <akitsch@google.com>
To keep a persistent backend storage for configuration, we will have to
keep a single source of truth. This involves supporting bi-directional
conversion between ToolsetConfig and Toolset.
This PR make the following changes:
* Embed ToolsetConfig in Toolset
* Add `ToConfig()` to extract ToolsetConfig from Toolset.
To keep a persistent backend storage for configuration, we will have to
keep a single source of truth. This involves supporting bi-directional
conversion between Config and AuthService.
This PR make the following changes:
* Embed Config in AuthService
* Add `ToConfig()` to extract Config from AuthService.
To keep a persistent backend storage for configuration, we will have to
keep a single source of truth. This involves supporting bi-directional
conversion between Config and Tool.
This PR make the following changes:
* Embed Config in Tool
* Add `ToConfig()` to extract Config from Tool.
Jules PR
---
*PR created automatically by Jules for task
[11947649751737965380](https://jules.google.com/task/11947649751737965380)*
---------
Co-authored-by: google-labs-jules[bot] <161369871+google-labs-jules[bot]@users.noreply.github.com>
Co-authored-by: Yuan Teoh <yuanteoh@google.com>
To keep a persistent backend storage for configuration, we will have to
keep a single source of truth. This involves supporting bi-directional
conversion between Config and Source.
This PR make the following changes:
* Embed Config in Source
* Add `ToConfig()` to extract Config from Source.
To facilitate the transition of moving invocation implementation to
Source, we will have to move parameter to `internal/util`. This approach
is crucial because certain parameters may not be fully resolvable
pre-implementation. Since both `internal/sources` and `internal/tools`
will need access to `parameters`, it will be more relevant to move
parameters implementation to utils.
## Description
> Should include a concise description of the changes (bug or feature),
it's
> impact, along with a summary of the solution
This PR adds a new tool to the Looker MCP Toolbox, that enables the user
authenticated to the Looker Source to generate authenticated embed urls
for dashboards, looks and queries. When combined with other tools that
already exist in the Looker toolbox, this enables searching existing
content and providing authenticated urls to them OR creating
authenticated urls from queries generated via Natural Language. The
embed urls will create an embed session for the user and can be opened
directly or added to an iframe `src` attribute to provide a smooth
Embedded Analytics setup.
Additionally this PR adds a new optional parameter to the Looker source
called `SessionLength` which an admin setting up the Looker source can
set to determine how long the Embed sessions last for.
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [ X] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [ X] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [ ] Ensure the tests and linter pass
- [ X] Code coverage does not decrease (if any source code was changed)
- [X ] Appropriate docs were updated (if necessary)
- [ X] Make sure to add `!` if this involve a breaking change
🛠️Fixes#1876https://github.com/googleapis/genai-toolbox/issues/1876
---------
Co-authored-by: Luka Fontanilla <maluka@google.com>
Co-authored-by: Dr. Strangelove <drstrangelove@google.com>
## Description
This pull request addresses issue
[#29](https://github.com/gemini-cli-extensions/mysql/issues/29) by
implementing default connection values for the MySQL data source. When a
user does not specify a host or port, the connection will now
automatically default to localhost:3306.
### Detailed Changes
1. **internal/prebuiltconfigs/tools/mysql.yaml**
- Added the default values to host and port.Specifically, they are now
host: ${MYSQL_HOST:localhost} and port: ${MYSQL_PORT:3306}.This allows
the configuration loader to directly use "localhost" or "3306" as a
fallback if the MYSQL_HOST or MYSQL_PORT environment variables are not
set.
**Testing and Validation**
To ensure the changes work correctly and are compatible with MySQL-like
databases, the feature was manually tested against a MariaDB instance
running in a Docker container.
The testing process involved the following steps:
1. A local testdb database with a products table was created in the
MariaDB container.
2. For Gemini CLI integration, the ~/.gemini/settings.json file was
configured to point to a local build of the toolbox executable.
3. The toolbox was launched in UI mode using the **_./toolbox
--prebuilt=mysql --ui_** command.
4. Two connection scenarios were tested to validate the new logic.
**Test 1: Default Connection (Host/Port Unset)**
The MYSQL_HOST and MYSQL_PORT environment variables were unset. The
application correctly defaulted to localhost:3306 and successfully
connected to the testdb database.
<img width="2087" height="1102" alt="Screenshot 2025-11-11 11 31 45 PM"
src="https://github.com/user-attachments/assets/a8ef5f92-eaca-472f-a0df-e2b3c0c027da"
/>
**Test 2: Explicit Connection (Host/Port Set)**
The MYSQL_HOST and MYSQL_PORT environment variables were explicitly set.
The application correctly used these values, overriding the defaults and
establishing a successful connection.
<img width="2073" height="958" alt="Screenshot 2025-11-11 6 12 44 PM"
src="https://github.com/user-attachments/assets/4b9b8838-091f-4c78-9e3b-97768323693c"
/>
**Result:**
In both scenarios, the list_tables prebuilt tool was executed via the
Toolbox UI, which successfully returned the products table from the
testdb database, confirming the changes work as expected.
Screenshot of list_tables execution from the Toolbox UI:
<img width="2251" height="1240" alt="result"
src="https://github.com/user-attachments/assets/f1c5372d-acc0-4551-af2d-fa1ee4b228d7"
/>
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [ ] Ensure the tests and linter pass
- [ ] Code coverage does not decrease (if any source code was changed)
- [ ] Appropriate docs were updated (if necessary)
- [ ] Make sure to add `!` if this involve a breaking change
🛠️ Fixes [#29](https://github.com/gemini-cli-extensions/mysql/issues/29)
---------
Co-authored-by: prernakakkar-google <158031829+prernakakkar-google@users.noreply.github.com>
### Description
This pull request addresses issue #41 by implementing default connection
values for the Postgresql data source. When a user does not specify a
host or port, the connection will now automatically default to
`localhost:5432`.
---
### Detailed Changes
- **docs/en/reference/prebuilt-tools.md**
-updated the env variables host and port to optional in the doc
- **internal/prebuiltconfigs/tools/postgres.yaml**
- Added the default values to host and port. Specifically, they are now
`host: ${POSTGRES_HOST:localhost}` and `port: ${POSTGRES_PORT:5432}`.
This allows the configuration loader to directly use `"localhost"` or
`"5432"` as a fallback if the `POSTGRES_HOST` or `POSTGRES_PORT`
environment variables are not set.
---
### PR Checklist
Thank you for opening a Pull Request! Before submitting your PR, there
are a few things you can do to make sure it goes smoothly:
- [ ] Make sure you reviewed
CONTRIBUTING.md
- [ ] Make sure to open an issue as a
bug/issue
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [ ] Ensure the tests and linter pass
- [ ] Code coverage does not decrease (if any source code was changed)
- [ ] Appropriate docs were updated (if necessary)
- [ ] Make sure to add ! if this involve a breaking change
🛠️ **Fixes** #41
---------
Co-authored-by: Averi Kitsch <akitsch@google.com>
## Description
---
This change introduces the `DefaultProject` field for the
`alloydb-admin` and `cloud-sql-admin` sources. This field allows the
alloydb and cloud sql control plane tools to use the project value from
the environment variables (Ex: `ALLOYDB_POSTGRES_PROJECT`) if it is
already set instead of asking the user.
## PR Checklist
---
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [x] Appropriate docs were updated (if necessary)
- [x] Make sure to add `!` if this involve a breaking change
🛠️ Fixes: https://github.com/gemini-cli-extensions/alloydb/issues/47
---------
Co-authored-by: Averi Kitsch <akitsch@google.com>
## Description
Add support for healthcare source, tool and prebuilt config. This branch
consist of all previously approved PRs.
🛠️Fixes#1648
---------
Co-authored-by: Marwan Tammam <15021613+Quarz0@users.noreply.github.com>
Assigning job iterator value to an array instead of map to preserve
column order.
When assigning incoming values to map, sometimes the result is not in
order of statement. E.g. `SELECT id, name from ...` might turn into
`{"name": "name_value", "id":1}` rather than `{"id":1, "name":
"name_value"}`. Previously, during json marshaling, it will ALWAYS order
the map in alphabetical order. so that wasn't an issue.
With the implementation of `orderedmap` (#1852), the bigquery execute
sql tool will now preserves the column order during the marshaling
process. Due to this, bigquery's integration test is flaky and failed
when the map is reordered. This update assign incoming value as array
instead, preserving the actual order.
## Description
The run_dashboard tool will run the query associated with each tile of
the dashboard and return the full set of query results. It enables the
agent to answer questions like "Summarize this dashboard for me".
---------
Co-authored-by: Yuan Teoh <45984206+Yuan325@users.noreply.github.com>
## Description
---
- This PR adds SingleStore database source and tools. The code is mostly
based on MySQL source and tools, and it uses the same go-mysql driver.
- https://github.com/singlestore-labs/singlestoredb-dev-image can be
used to deploy a test SingleStore instance. In this PR the default port
is set to 3308 so the command would be
```docker run \
-d --name singlestoredb-dev \
-e ROOT_PASSWORD="YOUR SINGLESTORE ROOT PASSWORD" \
-p 3308:3306 ghcr.io/singlestore-labs/singlestoredb-dev:latest
```
## PR Checklist
---
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/langchain-google-alloydb-pg-python/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [x] Appropriate docs were updated (if necessary)
- [x] Make sure to add `!` if this involve a breaking change
🛠️ Fixes https://github.com/googleapis/genai-toolbox/issues/1348
---------
Co-authored-by: Wenxin Du <117315983+duwenxin99@users.noreply.github.com>
## Description
Add a read-only PostgreSQL custom list_schemas tool, that returns the
schemas present in the database excluding system and temporary schemas.
Returns the schema name, schema owner, grants, number of functions,
number of tables, and number of views within each schema.
<img width="1985" height="1043" alt="Screenshot 2025-10-20 at 7 45
45 PM"
src="https://github.com/user-attachments/assets/8c4f0bb8-587c-489a-8795-efa79e92b06f"
/>
<img width="3372" height="1694" alt="3NpZG7W6h3XGsM7"
src="https://github.com/user-attachments/assets/370b5440-cc48-4c4e-82ea-4fd508cbcf2b"
/>
> Should include a concise description of the changes (bug or feature),
it's
> impact, along with a summary of the solution
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [x] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [x] Appropriate docs were updated (if necessary)
- [x] Make sure to add `!` if this involve a breaking change
🛠️ Fixes #<issue_number_goes_here>
Co-authored-by: Yuan Teoh <45984206+Yuan325@users.noreply.github.com>
Update bigquery test to include column order for SELECT statement.
Update mindsdb tests to drop table before creating. The whole
integration test pause when there's a failure from any one of
integration tests. If the test pause after `CREATE` and before `DROP`,
the int test will fail when running it again.
This commit introduces a new `orderedmap` package to preserve the column
order of SQL query results when they are marshaled to JSON.
The default Go `json.Marshal` function sorts map keys, which was causing
the column order to be lost in the output of the database tools.
This commit updates the following tools to use the new `orderedmap`
package:
- `mysqlexecutesql`
- `mssqlexecutesql`
- `postgresexecutesql`
- `spannerexecutesql`
- `sqliteexecutesql`
- `bigqueryexecutesql`
A new test has been added to the `mysqlexecutesql` tool to verify that
the column order is preserved.
## Description
> Should include a concise description of the changes (bug or feature),
it's
> impact, along with a summary of the solution
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [ ] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [ ] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [ ] Ensure the tests and linter pass
- [ ] Code coverage does not decrease (if any source code was changed)
- [ ] Appropriate docs were updated (if necessary)
- [ ] Make sure to add `!` if this involve a breaking change
🛠️Fixes#1492
---------
Co-authored-by: google-labs-jules[bot] <161369871+google-labs-jules[bot]@users.noreply.github.com>
Co-authored-by: Yuan Teoh <yuanteoh@google.com>
Co-authored-by: Yuan Teoh <45984206+Yuan325@users.noreply.github.com>
## Description
Corrects an issue where the `cloud-monitoring-query-prometheus` tool
would fail to populate the `authRequired` field in its generated
manifest.
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [ ] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [x] Ensure the tests and linter pass
- [x] Code coverage does not decrease (if any source code was changed)
- [ ] Appropriate docs were updated (if necessary)
- [ ] Make sure to add `!` if this involve a breaking change
🛠️ Fixes #<issue_number_goes_here>
🚀 Add MindsDB Integration: Expand Toolbox to Hundreds of Datasources
Overview
This PR introduces comprehensive MindsDB integration to the Google GenAI
Toolbox, enabling SQL queries across hundreds of datasources through a
unified interface. MindsDB is the most widely adopted AI federated
database that automatically translates MySQL queries into REST APIs,
GraphQL, and native protocols.
🎯 Key Value for Google GenAI Toolbox Ecosystem
1. Massive Datasource Expansion
Before: Toolbox limited to ~15 traditional databases
After: Access to hundreds of datasources including Salesforce, Jira,
GitHub, MongoDB, Gmail, Slack, and more
Impact: Dramatically expands the toolbox's reach and utility for
enterprise users
2. Cross-Datasource Analytics
New Capability: Perform joins and analytics across different datasources
seamlessly
Example: Join Salesforce opportunities with GitHub activity to correlate
sales with development activity
Value: Enables comprehensive data analysis that was previously
impossible
3. API Abstraction Layer
Innovation: Write standard SQL queries that automatically translate to
any API
Benefit: Developers can query REST APIs, GraphQL, and native protocols
using familiar SQL syntax
Impact: Reduces complexity and learning curve for accessing diverse
datasources
4. ML Model Integration
Enhanced Capability: Use ML models as virtual tables for real-time
predictions
Example: Query customer churn predictions directly through SQL
Value: Brings AI/ML capabilities into the standard SQL workflow
🔧 Technical Implementation
Source Layer
✅ New MindsDB source implementation using MySQL wire protocol
✅ Comprehensive test coverage with integration tests
✅ Updated existing MySQL tools to support MindsDB sources
✅ Created dedicated MindsDB tools for enhanced functionality
Tools Layer
✅ mindsdb-execute-sql: Direct SQL execution across federated datasources
✅ mindsdb-sql: Parameterized SQL queries with template support
✅ Backward compatibility with existing MySQL tools
Documentation & Configuration
✅ Comprehensive documentation with real-world examples
✅ Prebuilt configuration for easy setup
✅ Updated CLI help text and command-line options
📊 Supported Datasources
Business Applications
Salesforce (leads, opportunities, accounts)
Jira (issues, projects, workflows)
GitHub (repositories, commits, PRs)
Slack (channels, messages, teams)
HubSpot (contacts, companies, deals)
Databases & Storage
MongoDB (NoSQL collections as structured tables)
Redis (key-value stores)
Elasticsearch (search and analytics)
S3, filesystems, etc (file storage)
Communication & Email
Gmail/Outlook (emails, attachments)
Microsoft Teams (communications, files)
Discord (server data, messages)
🎯 Example Use Cases
Cross-Datasource Analytics
-- Join Salesforce opportunities with GitHub activity
```
SELECT
s.opportunity_name,
s.amount,
g.repository_name,
COUNT(g.commits) as commit_count
FROM salesforce.opportunities s
JOIN github.repositories g ON s.account_id = g.owner_id
WHERE s.stage = 'Closed Won';
```
Email & Communication Analysis
```
-- Analyze email patterns with Slack activity
SELECT
e.sender,
e.subject,
s.channel_name,
COUNT(s.messages) as message_count
FROM gmail.emails e
JOIN slack.messages s ON e.sender = s.user_name
WHERE e.date >= '2024-01-01';
```
🚀 Benefits for Google GenAI Toolbox
Enterprise Adoption: Enables access to enterprise datasources
(Salesforce, Jira, etc.)
Developer Productivity: Familiar SQL interface for any datasource
AI/ML Integration: Seamless integration of ML models into SQL workflows
Scalability: Single interface for hundreds of datasources
Competitive Advantage: Unique federated database capabilities in the
toolbox ecosystem
📈 Impact Metrics
Datasource Coverage: +1000% increase in supported datasources
API Abstraction: Eliminates need to learn individual API syntaxes
Cross-Platform Analytics: Enables previously impossible data
correlations
ML Integration: Brings AI capabilities into standard SQL workflows
🔗 Resources
MindsDB Documentation
MindsDB GitHub
Updated Toolbox Documentation
✅ Testing
✅ Unit tests for MindsDB source implementation
✅ Integration tests with real datasource examples
✅ Backward compatibility with existing MySQL tools
✅ Documentation examples tested and verified
This integration transforms the Google GenAI Toolbox from a traditional
database tool into a comprehensive federated data platform, enabling
users to query and analyze data across their entire technology stack
through a unified SQL interface.
---------
Co-authored-by: duwenxin <duwenxin@google.com>
Co-authored-by: setohe0909 <setohe.09@gmail.com>
Co-authored-by: Kurtis Van Gent <31518063+kurtisvg@users.noreply.github.com>
Co-authored-by: Wenxin Du <117315983+duwenxin99@users.noreply.github.com>
Co-authored-by: Yuan Teoh <45984206+Yuan325@users.noreply.github.com>
## Description
The order of parameters in alloydb_ai_nl.execute_nl_query changed, which
broke the alloydbainl tool. This adds named parameters to the statement
in the tool, which fixes this.
This will be a breaking change for existing user that defined their
natural language configuration with the `create_configure` operation.
The `execute_nl_query` input argument parameter and order had been
updated recently, hence, users that are defining their configuration
with the latest instructions will not be able to use Toolbox. This
update is inevitable.
Previously, user will create the configuration with the following:
```
CALL google_ml.create_model(model_id => 'gemini-2_0_flash', ...);
SELECT alloydb_ai_nl.g_manage_configuration(
'create_configuration', -- operation
'my_nl_config', -- configuration_id_in
'gemini-2_0_flash' -- model_id_in
);
SELECT alloydb_ai_nl.g_manage_configuration(
operation => 'register_table_view',
configuration_id_in => 'my_nl_config',
table_views_in=>'{auth_psv}');
```
Currently, user will create the configuration with the following:
```
SELECT alloydb_ai_nl.g_create_configuration(configuration_id =>'my_nl_config');
SELECT alloydb_ai_nl.g_manage_configuration(
operation => 'register_table_view',
configuration_id_in => 'my_nl_config',
table_views_in=>'{auth_psv}'
);
```
This PR also updates the nl_question from "return 1" to "return the
number 1" to provide more context to the model.
A new `ainl_update_testing` database was created with the new NL
configuration in order to not break existing integration tests before
merging this PR. Once this is merged, the existing `test_database`
database will be updated and will update the integration test's database
again.
## PR Checklist
> Thank you for opening a Pull Request! Before submitting your PR, there
are a
> few things you can do to make sure it goes smoothly:
- [X] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [X] Make sure to open an issue as a
[bug/issue](https://github.com/googleapis/genai-toolbox/issues/new/choose)
before writing your code! That way we can discuss the change, evaluate
designs, and agree on the general idea
- [X] Ensure the tests and linter pass
- [X] Code coverage does not decrease (if any source code was changed)
- [X] Appropriate docs were updated (if necessary)
- [X] Make sure to add `!` if this involve a breaking change
🛠️Fixes#1752
---------
Co-authored-by: Yuan Teoh <yuanteoh@google.com>
Co-authored-by: Yuan Teoh <45984206+Yuan325@users.noreply.github.com>
