Fix previously failing test. During unmarshal, if the resources is not
presented, return nil instead of empty struct.
Also added a note in docs to view the configuration file v1 version.
Update yaml tag in tools from `kind` to `type`.
Upcoming PRs:
* Update config in `/tests` and `/docs`
Will resolve other unit tests failure in the final PR to keep things
simpler.
Update `parseToolsFile()` to use the file conversion function. Update
unmarshalling function since we will be unmarshalling per-doc (separated
by `---`).
`AuthSources` will be converted to `AuthServices` during the file
conversion stage. Hence, removing it from `ToolsFile` struct. ToolsFile
v2 will not support `AuthSources`. Double checked that all docs do not
reference to `AuthSources` anymore; new users is expected to be using
`AuthServices` instead.
This PR will not pass the unit test since the updates for resources's
yaml tag will be in subsequent PRs. Breaking it down to keep review
simpler.
Related #817
Upcoming PR:
* Update yaml tag
Add preprocessing function during parsing of tools file to convert v2
tools file.
Related #817
Future updates will include:
* Update unmarshal function for ToolsFile to convert config file (test
will fail since the yaml tag is not yet updated).
* Update yaml tag (test will pass).
---------
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Averi Kitsch <akitsch@google.com>
Update source code `Kind` to `Type`. It's only changes within our code.
Changes to yaml tag (that will affect users) will be done in later PRs.
This is a breaking change since it updates telemetry's span attribute
from `source_kind` to `source_type`.
Related #817
Future updates will include:
* Updating a preprocessing function to convert config file from v1 to v2
* Update unmarshal function for ToolsFile to convert config file (test
will fail since the yaml tag is not yet updated).
* Update yaml tag (test will pass).
### Description
fix: Surface Dataplex API errors in MCP results
This change addresses issue
https://github.com/googleapis/genai-toolbox/issues/2203, where Dataplex
API errors, such as '403 Forbidden' (Permission Denied), were not being
properly surfaced in the MCP tool results. Previously, these critical
API errors would manifest as generic "connection interrupted" messages,
significantly hindering developer debugging and trust in the Toolbox.
The fix enhances the error handling within the 'dataplexsearchentries'
and 'dataplexsearchaspecttypes' tools. When an error occurs during the
iteration of Dataplex API results, the system now:
Utilizes 'google.golang.org/grpc/status.FromError' to attempt to convert
the returned error into a gRPC status. This is crucial because Google
Cloud client libraries often return errors compatible with gRPC.
If the error is a gRPC status, the canonical error code (e.g.,
'codes.PermissionDenied') and the associated error message are
extracted.
This ensures that users receive clear actionable error feedback,
allowing for quicker diagnosis and resolution of issues like missing IAM
permissions. This aligns with best practices for API error surfacing,
improving the usability and reliability of the Dataplex tools within the
GenAI Toolbox.
Fixes https://github.com/googleapis/genai-toolbox/issues/2203
## 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>
---------
Co-authored-by: Wenxin Du <117315983+duwenxin99@users.noreply.github.com>
Co-authored-by: Averi Kitsch <akitsch@google.com>
oceanbase's link is hitting `Rejected status code (this depends on your
"accept" configuration): Too Many Requests` error. It might have
temporarily blocked Lychee due to the too many request being sent in a
short time. Adding the link to lycheeignore to unblock GHA failure.
## Description
This PR introduces support for merging multiple prebuilt configurations.
To ensure compatibility, the following restrictions apply:
- No Naming Collisions: Configurations cannot share duplicate names for
any resources (Tools, Sources, Toolsets, Auth Services, etc.).
- Shared Environment Variables: If multiple sources rely on the same
environment variable, they must share the same value; unique values for
the same variable are not supported
## Usage Examples
### Successful Initialization
You can load multiple prebuilt configurations by either repeating the
--prebuilt flag or by providing a comma-separated list.
**Option 1:** Multiple Flags
```
./toolbox --prebuilt alloydb-postgres --prebuilt alloydb-postgres-admin
```
**Option 2:** Comma-Separated Values
```
./toolbox --prebuilt alloydb-postgres,alloydb-postgres-admin
```
### Initialization Failure (Resource Conflict)
If two or more configurations define a resource with the same name (such
as a Tool or Source, etc.), the server will fail to start and display a
conflict error.
```
./toolbox --prebuilt alloydb-postgres --prebuilt cloud-sql-mysql
2026-01-13T11:14:50.758121799Z INFO "Using prebuilt tool configurations for: alloydb-postgres, cloud-sql-mysql"
2026-01-13T11:14:50.764578167Z ERROR "resource conflicts detected:\n - tool 'execute_sql' (file #2)\n - tool 'list_active_queries' (file #2)\n - tool 'get_query_plan' (file #2)\n - tool 'list_tables' (file #2)\n\nPlease ensure each source, authService, tool, toolset and prompt has a unique name across all files"
```
## 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#1855
---------
Co-authored-by: Averi Kitsch <akitsch@google.com>
Check for naming validation for Tool. This validation follows the MCP
SEP986 [naming
guidance](1b1eb60ec4/docs/specification/draft/server/tools.mdx (tool-names)).
Name will be checked before MCP initialization (where specs version is
confirmed). Hence, we will be implementing this across all versions and
endpoints.
This will be a breaking change for user that currently uses other
special character as name (other than `_`, `-`, `.`)
This PR introduces a new configuration field valueFromParam to the tool
definitions. This feature allows a parameter to automatically inherit
its value from another sibling parameter, mainly to streamline the
configuration of vector insertion tools.
Parameters utilizing valueFromParam are excluded from the Tool and MCP
manifests. This means the LLM does not see these parameters and is not
required to generate them. The value is resolved internally by the
Toolbox during execution.
## Description
Use t.Cleanup() to register cleanup of FHIR and DICOM stores immediately
after creation. This fixes the uncleaned FHIR/DICOM stores that remain
in the project(In the earlier implementation, teardown does not get
triggered if the test failed).
🛠️Fixes#1986
---------
Co-authored-by: Yuan Teoh <yuanteoh@google.com>
Co-authored-by: Yuan Teoh <45984206+Yuan325@users.noreply.github.com>
## Description
Add a new `--user-agent-metadata` flag that allows user to append
additional user agent metadata. The flag takes in []string and will
concatenate it with `.`.
```
go run . --user-agent-metadata=foo
```
produces `0.25.0+dev.darwin.arm64+foo` user agent string
```
go run . --user-agent-metadata=foo,bar
```
produces `0.25.0+dev.darwin.arm64+foo+bar` user agent string
## 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>
Previous refactoring (#2273) accidentally removed the authorization
checks prior to token retrieval. This issue went unnoticed because the
integration tests were disabled. I am re-adding the necessary checks.
## Description
Update the GDA source document to clarify that only `AlloyDbReference`,
`SpannerReference`, and `CloudSqlReference` are supported.
## 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#2324
---------
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
## Description
Partially fixes
https://github.com/googleapis/mcp-toolbox-sdk-python/issues/496
## 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
- [ ] 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 pull request adds a new tool, cloud-sql-restore-backup, which
enables restoring a backup onto a Cloud SQL instance from the toolbox
using the Cloud SQL Admin API. The tool supports restoring standard,
project level, and BackupDR backups.
Tested:
<img width="3758" height="532" alt="image"
src="https://github.com/user-attachments/assets/d1d61af7-d96e-417c-898c-65b876de4c5e"
/>
## 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#2170
Co-authored-by: Averi Kitsch <akitsch@google.com>
Update error code from 400 to 403 according to MCP
[updates](https://github.com/modelcontextprotocol/modelcontextprotocol/pull/1439)
for invalid origin header.
Also updated hostCheck to only check host, not port.
To test, run Toolbox with the following (also work with port number e.g.
`--allowed-host=127.0.0.1:5000`):
```
go run . --allowed-hosts=127.0.0.1
```
Test with the following:
```
// curl successfully
curl -H "Host: 127.0.0.1:5000" http://127.0.0.1:5000
// curl successfully
curl -H "Host: 127.0.0.1:3000" http://127.0.0.1:5000
// will show Invalid Host Header error
curl -H "Host: attacker:5000" http://127.0.0.1:5000
```
Add remaining CLI flags for the server published on official mcp
registry.
ref: https://googleapis.github.io/genai-toolbox/reference/cli/
_note: mcp registry do not support shorthand flag (there are no options
to defined an alternate name). The only way is to define them as
separate named arguments but it may not work well since both would try
to set the same underlying value._
- Refactors the 'cloud-gemini-data-analytics-query' tool to update its
default description with detailed tool guidance and usage guidance.
- Append the default description to the tools.yaml description no matter
whether the tools.yaml description exists since this guidance will
always be useful to the agent on how to use the tool.
- Renames the input parameter from 'prompt' to 'query' for better
consistency.
## Description
This pull request adds a new tool, cloud-sql-create-backup, which
enables taking a backup on a Cloud SQL instance from the toolbox using
the Cloud SQL Admin API. The tool supports optionally supplying a
location or description for the backup.
Tested:
<img width="1561" height="425" alt="image"
src="https://github.com/user-attachments/assets/c8984b07-5450-470a-9ac6-df16943e25e9"
/>
## 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#2140
---------
Co-authored-by: Yuan Teoh <45984206+Yuan325@users.noreply.github.com>
Co-authored-by: Averi Kitsch <akitsch@google.com>
## Description
Samples for MCP with Neo4j for this page:
https://googleapis.github.io/genai-toolbox/samples/
## 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)
- [x] Appropriate docs were updated (if necessary)
- [x] Make sure to add `!` if this involve a breaking change
---------
Co-authored-by: Yuan Teoh <45984206+Yuan325@users.noreply.github.com>
## Description
Addresses an issue where Dataplex AspectTypes created during integration
tests were not consistently deleted. This accumulation led to exceeding
the AspectType quota.
To prevent this, the test setup now includes a step to list and delete
all existing AspectTypes within the test project and location *before*
attempting to create any new ones. This ensures a clean state for each
test run and avoids hitting the quota.
## 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#2057
Co-authored-by: Averi Kitsch <akitsch@google.com>
Update existing docs to install PSV extensions **only** when needed.
This tool could be use without installing PSV if NLConfigParam is not
used.
---------
Co-authored-by: Averi Kitsch <akitsch@google.com>
Move source-related queries from `Invoke()` function into Source.
This is an effort to generalizing tools to work with any Source that
implements a specific interface. This will provide a better segregation
of the roles for Tools vs Source.
Tool's role will be limited to the following:
* Resolve any pre-implementation steps or parameters (e.g. template
parameters)
* Retrieving Source
* Calling the source's implementation
Move source-related queries from `Invoke()` function into Source.
This is an effort to generalizing tools to work with any Source that
implements a specific interface. This will provide a better segregation
of the roles for Tools vs Source.
Tool's role will be limited to the following:
* Resolve any pre-implementation steps or parameters (e.g. template
parameters)
* Retrieving Source
* Calling the source's implementation
## 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:
- [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
- [ ] 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>
Co-authored-by: Averi Kitsch <akitsch@google.com>
Move source-related queries from `Invoke()` function into Source.
This is an effort to generalizing tools to work with any Source that
implements a specific interface. This will provide a better segregation
of the roles for Tools vs Source.
Tool's role will be limited to the following:
* Resolve any pre-implementation steps or parameters (e.g. template
parameters)
* Retrieving Source
* Calling the source's implementation
Move source-related queries from `Invoke()` function into Source.
This PR addresses the following sources:
* dataplex
* http
* serverlessspark
This is an effort to generalizing tools to work with any Source that
implements a specific interface. This will provide a better segregation
of the roles for Tools vs Source.
Tool's role will be limited to the following:
* Resolve any pre-implementation steps or parameters (e.g. template
parameters)
* Retrieving Source
* Calling the source's implementation
Move source-related queries from `Invoke()` function into Source.
This is an effort to generalizing tools to work with any Source that
implements a specific interface. This will provide a better segregation
of the roles for Tools vs Source.
Tool's role will be limited to the following:
* Resolve any pre-implementation steps or parameters (e.g. template
parameters)
* Retrieving Source
* Calling the source's implementation
This change allows the agent developer to control the maxium number of
rows returned from tools running BigQuery SQL query. Using this feature
the agent developer could limit how large output is presented to LLM in
an agentic user journey.
## 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:
- [x] Make sure you reviewed
[CONTRIBUTING.md](https://github.com/googleapis/genai-toolbox/blob/main/CONTRIBUTING.md)
- [ ] Make sure to open an issue
https://github.com/googleapis/genai-toolbox/issues/2261
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#2261 2261
## Description
update docs to reflect triaging workflow and SLO
## 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>
## Description
PSV should only be required when when it is needed. Currently, we
require psv even whenever user uses AlloyDB AI NL tool. This is due to
the statement that we use to execute nl query.
This PR modified the statement query to only utilize `param_names` and
`param_values` when needed.
Manually tested with a db that does not have psv installed.
🛠️Fixes#1970
## Description
Add default value to manifest (for both native endpoint and mcp
endpoint).
## 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#1602
