feat: add LlamaIndex TS support to JS SDK #1178

abhishekpatil4 · 2025-01-10T18:31:20Z

Added support for LlamaIndex TS framework in JS SDK
Created test cases for framework validation

Important

Add LlamaIndex TypeScript framework support to JavaScript SDK with new class, tests, and dependencies.

Framework Support:
- Added LlamaIndexToolSet class in llamaindex.ts for LlamaIndex framework support.
- Integrated LlamaIndexToolSet into index.ts for export.
Testing:
- Created llamaindex.spec.ts with test cases for LlamaIndexToolSet.
Dependencies:
- Added @llamaindex/core and llamaindex to package.json dependencies and devDependencies.
Misc:
- Minor formatting fix in smol_demo.py.

^{This description was created by}^{for fb89115. It will automatically update as commits are pushed.}

vercel · 2025-01-10T18:31:26Z

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name	Status	Preview	Comments	Updated (UTC)
composio	✅ Ready (Inspect)	Visit Preview	💬 Add feedback	Feb 11, 2025 7:17am

entelligence-ai-pr-reviews · 2025-01-10T18:32:01Z

js/src/frameworks/llamaindex.ts

+        parameters: {
+          type: "object",
+          properties: schema.parameters.properties || {},
+          required: schema.parameters.required || [],


The getTools function in llamaindex.ts does not correctly handle the case where schema.parameters is undefined. This can lead to a runtime error when trying to access properties or required.

entelligence-ai-pr-reviews · 2025-01-10T18:32:02Z

js/src/frameworks/llamaindex.spec.ts

+        owner: "composioHQ",
+        repo: "composio",
+      },
+      entityId: "default",
+      connectedAccountId: "9442cab3-d54f-4903-976c-ee67ef506c9b",


The test Should create custom action to star a repository in llamaindex.spec.ts uses hardcoded values for connectedAccountId and the repository details. This makes the test less robust and prone to failures if these values change.

entelligence-ai-pr-reviews · 2025-01-10T18:32:03Z

js/src/frameworks/llamaindex.spec.ts

+    await llamaindexToolSet.createAction({
+      actionName: "starRepositoryCustomAction",
+      toolName: "github",
+      description: "This action stars a repository",
+      inputParams: z.object({
+        owner: z.string(),
+        repo: z.string(),
+      }),
+      callback: async (
+        inputParams,
+        _authCredentials,
+        executeRequest
+      ): Promise<ActionExecuteResponse> => {
+        const res = await executeRequest({
+          endpoint: `/user/starred/${inputParams.owner}/${inputParams.repo}`,
+          method: "PUT",
+          parameters: [],
+        });
+        return res;
+      },
+    });


The createAction call in the test Should create custom action to star a repository lacks proper error handling. If the action creation fails, the test will not fail and the subsequent getTools and executeAction calls might behave unexpectedly.

github-actions · 2025-01-10T18:32:04Z

This comment was generated by github-actions[bot]!

JS SDK Coverage Report

📊 Coverage report for JS SDK can be found at the following URL:
https://pub-92e668239ab84bfd80ee07d61e9d2f40.r2.dev/coverage-13257725668/coverage/index.html

📁 Test report folder can be found at the following URL:
https://pub-92e668239ab84bfd80ee07d61e9d2f40.r2.dev/html-report-13257725668/html-report/report.html

entelligence-ai-pr-reviews · 2025-01-10T18:32:05Z

🚀 Code review completed!

ellipsis-dev

👍 Looks good to me! Reviewed everything up to b7a8481 in 58 seconds

More details

Looked at 192 lines of code in 3 files
Skipped 0 files when reviewing.
Skipped posting 4 drafted comments based on config settings.

1. js/src/frameworks/llamaindex.spec.ts:16

Draft comment:
The test name 'getools' is not descriptive. Consider renaming it to something more meaningful, like 'should return tools array'.
Reason this comment was not posted:
Confidence changes required: 50%
The test name 'getools' is not descriptive and seems to be a typo. It should be more descriptive to reflect the test's purpose.

2. js/src/frameworks/llamaindex.spec.ts:22

Draft comment:
The test case 'check if tools are coming' is redundant with 'check if getTools, actions are coming'. Consider removing one to avoid duplication.
Reason this comment was not posted:
Confidence changes required: 50%
The test case 'check if tools are coming' and 'check if getTools, actions are coming' are redundant as they perform the same check. One of them should be removed to avoid duplication.

3. js/src/frameworks/llamaindex.spec.ts:63

Draft comment:
Typo in variable name 'actionOuput'. It should be 'actionOutput'.
Reason this comment was not posted:
Confidence changes required: 50%
The variable name 'actionOuput' is misspelled. It should be 'actionOutput'.

4. js/src/frameworks/llamaindex.ts:36

Draft comment:
Avoid using 'any' type for 'schema'. Use a more specific type to ensure type safety.
Reason this comment was not posted:
Confidence changes required: 50%
The use of 'any' type in TypeScript should be avoided as it defeats the purpose of type safety. A more specific type should be used for 'schema'.

Workflow ID: wflow_1ImRrec93dw0lj3s

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

shreysingla11 · 2025-01-10T18:33:07Z

js/src/frameworks/llamaindex.spec.ts

+import { ActionExecuteResponse } from "../sdk/models/actions";
+import { LlamaIndexToolSet } from "./llamaindex";
+
+describe("Apps class tests", () => {


The test suite description Apps class tests is misleading as it's testing the LlamaIndexToolSet class. Consider renaming it to LlamaIndexToolSet tests for better clarity.

shreysingla11 · 2025-01-10T18:33:16Z

js/src/frameworks/llamaindex.ts

+    });
+  }
+
+  private _wrapTool(


The _wrapTool method accepts schema: any as a parameter type. Consider creating a proper interface for better type safety:

interface ToolSchema { name: string; description: string; parameters: { properties?: Record<string, unknown>; required?: string[]; }; }

shreysingla11 · 2025-01-10T18:33:23Z

js/src/frameworks/llamaindex.ts

+    entityId: Optional<string> = null
+  ): FunctionTool<Record<string, unknown>, JSONValue | Promise<JSONValue>> {
+    return FunctionTool.from(
+      async (params: Record<string, unknown>) => {


Consider adding error handling for the JSON parsing operation:

try { return JSON.parse(JSON.stringify(result)) as JSONValue; } catch (error) { throw new Error(`Failed to parse tool result: ${error.message}`); }

This would provide better error messages and make debugging easier.

shreysingla11 · 2025-01-10T18:33:35Z

Code Review Summary

Overall Assessment

The implementation of LlamaIndex TypeScript framework support is well-structured and follows the established patterns in the codebase. The code is generally clean and functional, with good test coverage.

Positive Aspects

Good integration with the base ComposioToolSet class
Proper implementation of LlamaIndex's FunctionTool interface
Clean removal of unused OpenAI-specific code
Good test coverage for basic functionality

Suggestions for Improvement

Type Safety
- Replace any type in _wrapTool with a proper interface
- Add proper error handling for JSON operations
Testing
- Rename test suite to accurately reflect what's being tested
- Consider adding error case tests
- Add tests for edge cases with empty or invalid filters
Documentation
- Add JSDoc comments for public methods
- Consider adding examples in comments for common use cases

Code Quality Rating: 8/10

The code is well-implemented and follows good practices, with minor improvements suggested for type safety and documentation.

plxity · 2025-01-10T18:33:49Z

@abhishekpatil4 Couple of actions are failing, please fix them

entelligence-ai-pr-reviews · 2025-01-10T19:02:06Z

js/src/frameworks/llamaindex.ts

+        parameters: {
+          type: "object",
+          properties: schema.parameters.properties || {},
+          required: schema.parameters.required || [],
+        },


The getTools function in llamaindex.ts does not correctly handle the case when schema.parameters is null or undefined, which could lead to a runtime error.

entelligence-ai-pr-reviews · 2025-01-10T19:02:07Z

js/src/frameworks/llamaindex.spec.ts

+  it("getools", async () => {
+    const tools = await llamaindexToolSet.getTools({
+      apps: ["github"],
+    });
+    expect(tools).toBeInstanceOf(Array);
+  });


The test getools in llamaindex.spec.ts doesn't assert anything meaningful. It only checks the type of the return, not the content.

entelligence-ai-pr-reviews · 2025-01-10T19:02:08Z

js/src/frameworks/llamaindex.spec.ts

+  it("Should create custom action to star a repository", async () => {
+    await llamaindexToolSet.createAction({
+      actionName: "starRepositoryCustomAction",
+      toolName: "github",
+      description: "This action stars a repository",
+      inputParams: z.object({
+        owner: z.string(),
+        repo: z.string(),
+      }),
+      callback: async (
+        inputParams,
+        _authCredentials,
+        executeRequest
+      ): Promise<ActionExecuteResponse> => {
+        const res = await executeRequest({
+          endpoint: `/user/starred/${inputParams.owner}/${inputParams.repo}`,
+          method: "PUT",
+          parameters: [],
+        });
+        return res;
+      },
+    });
+
+    const tools = await llamaindexToolSet.getTools({
+      actions: ["starRepositoryCustomAction"],
+    });
+
+    await expect(tools.length).toBe(1);
+    const actionOuput = await llamaindexToolSet.executeAction({
+      action: "starRepositoryCustomAction",
+      params: {
+        owner: "composioHQ",
+        repo: "composio",
+      },
+      entityId: "default",
+      connectedAccountId: "9442cab3-d54f-4903-976c-ee67ef506c9b",
+    });
+
+    expect(actionOuput).toHaveProperty("successful", true);
+  });


The test suite in llamaindex.spec.ts is missing a test to verify that the createAction method adds a new action correctly. The current test only checks if the number of tools increased but not the properties of the new tool/action.

ellipsis-dev

❌ Changes requested. Incremental review on 769477f in 59 seconds

More details

Looked at 26 lines of code in 1 files
Skipped 0 files when reviewing.
Skipped posting 0 drafted comments based on config settings.

Workflow ID: wflow_bml1QyLngGW09OCX

Want Ellipsis to fix these issues? Tag @ellipsis-dev in a comment. You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

ellipsis-dev · 2025-01-10T19:02:50Z

js/src/frameworks/llamaindex.ts

+type ToolSchema = {
+    name: string;
+    description: string;
+    parameters: Record<string, unknown>;


The parameters property in ToolSchema is defined as Record<string, unknown>, but it is accessed as if it has properties and required fields in _wrapTool. Consider defining parameters with a more specific type that includes properties and required fields to avoid potential runtime errors.

entelligence-ai-pr-reviews · 2025-01-10T19:06:38Z

js/src/frameworks/llamaindex.ts

+        parameters: {
+          type: "object",
+          properties: schema.parameters.properties || {},
+          required: schema.parameters.required || [],
+        },


The getTools function in llamaindex.ts does not correctly handle the case where parameters is undefined in the tool schema. This could lead to a runtime error when trying to access schema.parameters.properties or schema.parameters.required.

entelligence-ai-pr-reviews · 2025-01-10T19:06:39Z

js/src/frameworks/llamaindex.spec.ts

+  it("getools", async () => {
+    const tools = await llamaindexToolSet.getTools({
+      apps: ["github"],
+    });
+    expect(tools).toBeInstanceOf(Array);
+  });


The test getools in llamaindex.spec.ts doesn't assert anything meaningful. It only checks if the returned value is an array, but not its contents. This test should verify the expected tools are returned.

)

Co-authored-by: Himanshu Dixit <[email protected]>

#1207)

ellipsis-dev · 2025-01-20T12:55:45Z

⚠️ This PR is too big for Ellipsis, but support for larger PRs is coming soon. If you want us to prioritize this feature, let us know at [email protected]

Generated with ❤️ by ellipsis.dev

entelligence-ai-pr-reviews · 2025-02-03T09:52:35Z

js/src/frameworks/llamaindex.spec.ts

+    const connectedAccount = await llamaindexToolSet.connectedAccounts.list({
+      appNames: "github",
+      showActiveOnly: true,
+    });
+
+    const actionOuput = await llamaindexToolSet.executeAction({
+      action: "starRepositoryCustomAction",
+      params: {
+        owner: "composioHQ",
+        repo: "composio",
+      },
+      entityId: "default",
+      connectedAccountId: connectedAccount.items[0].id,


Missing error handling for empty connectedAccount.items array which could cause runtime error when accessing index 0. Should validate array length before accessing.

📝 Committable Code Suggestion

‼️ Ensure you review the code suggestion before committing it to the branch. Make sure it replaces the highlighted code, contains no missing lines, and has no issues with indentation.

Suggested change

const connectedAccount = await llamaindexToolSet.connectedAccounts.list({

appNames: "github",

showActiveOnly: true,

});

const actionOuput = await llamaindexToolSet.executeAction({

action: "starRepositoryCustomAction",

params: {

owner: "composioHQ",

repo: "composio",

},

entityId: "default",

connectedAccountId: connectedAccount.items[0].id,

const connectedAccount = await llamaindexToolSet.connectedAccounts.list({

appNames: "github",

showActiveOnly: true,

});

if (!connectedAccount.items?.length) {

throw new Error('No connected GitHub accounts found');

}

const actionOuput = await llamaindexToolSet.executeAction({

...,

connectedAccountId: connectedAccount.items[0].id,

});

ellipsis-dev

👍 Looks good to me! Incremental review on a08dfe4 in 42 seconds

More details

Looked at 26 lines of code in 1 files
Skipped 0 files when reviewing.
Skipped posting 4 drafted comments based on config settings.

1. js/src/frameworks/llamaindex.spec.ts:64

Draft comment:
Verify that 'connectedAccount.items' is non-empty before accessing index 0.
Reason this comment was not posted:
Comment did not seem useful.

2. js/src/frameworks/llamaindex.spec.ts:76

Draft comment:
Using connectedAccount.items[0].id may cause runtime errors if no accounts available.
Reason this comment was not posted:
Marked as duplicate.

3. js/src/frameworks/llamaindex.spec.ts:64

Draft comment:
Consider adding an assertion to ensure 'connectedAccount.items' is not empty before accessing [0].
Reason this comment was not posted:
Marked as duplicate.

4. js/src/frameworks/llamaindex.spec.ts:76

Draft comment:
Typo: 'actionOuput' should likely be 'actionOutput'.
Reason this comment was not posted:
Comment was on unchanged code.

Workflow ID: wflow_Y8xadwaOqIL7gzfT

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

ellipsis-dev

👍 Looks good to me! Incremental review on 45efb71 in 53 seconds

More details

Looked at 42 lines of code in 1 files
Skipped 1 files when reviewing.
Skipped posting 6 drafted comments based on config settings.

1. js/package.json:41

Draft comment:
Ensure @llamaindex/core version is consistent in both peerDependencies and devDependencies.
Reason this comment was not posted:
Comment did not seem useful.

2. js/package.json:55

Draft comment:
Verify 'resolve-package-path' is not duplicated and ordering is correct.
Reason this comment was not posted:
Comment did not seem useful.

3. js/package.json:111

Draft comment:
New packageManager field; confirm lock files (pnpm-lock.yaml) are updated accordingly.
Reason this comment was not posted:
Confidence changes required: 33%
None

4. js/package.json:41

Draft comment:
Added '@llamaindex/core'. Verify if adding it in peerDependencies and devDependencies is intentional.
Reason this comment was not posted:
Confidence changes required: 33%
None

5. js/package.json:55

Draft comment:
Removed duplicate 'resolve-package-path' entry; ensure its placement is correct.
Reason this comment was not posted:
Confidence changes required: 33%
None

6. js/package.json:107

Draft comment:
The added 'packageManager' field enforces pnpm consistency; this is a good practice.
Reason this comment was not posted:
Confidence changes required: 0%
None

Workflow ID: wflow_f4YKceYj0fziOR9n

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

ellipsis-dev

👍 Looks good to me! Incremental review on fb89115 in 22 seconds

More details

Looked at 12 lines of code in 1 files
Skipped 0 files when reviewing.
Skipped posting 2 drafted comments based on config settings.

1. python/plugins/smolagent/smol_demo.py:14

Draft comment:
Minor formatting change; inline comment spacing improved.
Reason this comment was not posted:
Confidence changes required: 0% <= threshold 50%
None

2. python/plugins/smolagent/smol_demo.py:14

Draft comment:
Minor whitespace fix for consistency in the type: ignore comment.
Reason this comment was not posted:
Confidence changes required: 0% <= threshold 50%
None

Workflow ID: wflow_YQL4UilqDKawL5Eg

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

abhishekpatil4 added 3 commits January 10, 2025 17:56

feat: add support for llamaindex (TS)

10895d3

fix: llamaindex tests

f33976d

feat: add LlamaIndex TS framework support

b7a8481

abhishekpatil4 requested review from himanshu-dixit, plxity and utkarsh-dixit as code owners January 10, 2025 18:31

entelligence-ai-pr-reviews bot reviewed Jan 10, 2025

View reviewed changes

ellipsis-dev bot reviewed Jan 10, 2025

View reviewed changes

abhishekpatil4 changed the title ~~Feat/llamaindex-ts~~ Fix: removed some methods Jan 10, 2025

shreysingla11 reviewed Jan 10, 2025

View reviewed changes

himanshu-dixit approved these changes Jan 10, 2025

View reviewed changes

fix: added type to schema

769477f

vercel bot had a problem deploying to Preview January 10, 2025 19:01 Failure

entelligence-ai-pr-reviews bot reviewed Jan 10, 2025

View reviewed changes

ellipsis-dev bot reviewed Jan 10, 2025

View reviewed changes

fix: lint

57535ff

vercel bot had a problem deploying to Preview January 10, 2025 19:06 Failure

entelligence-ai-pr-reviews bot reviewed Jan 10, 2025

View reviewed changes

himanshu-dixit and others added 9 commits January 20, 2025 18:23

feat: change TEST_ENVIRONMENT -> TEST_ENV and log base url for CI (#1198

c61edc4

)

feat: integration list, reinitiate and update schema (#1189)

0eb60c7

feat: Added helicone for caching openai response (#1202)

37c7054

fix: enforce specific NodeJS version (#943)

8647237

Co-authored-by: Himanshu Dixit <[email protected]>

feat: crypto kit agents added (#1205)

ee56d22

feat: triggerId -> triggerName, triggerInstanceId, add methods for co… (

9b5db7f

#1207)

fix: allow auto reconnection in pusher client (#1199)

bd7906e

fix: filter connections by entity [ENG-2523] (#1200)

058121e

fix: session ID delegation (#1212)

b08fcd1

abhishekpatil4 requested a review from angrybayblade as a code owner January 20, 2025 12:54

vercel bot deployed to Preview January 20, 2025 12:55 View deployment

Merge branch 'master' into feat/llamaindex-ts

2cc4b51

vercel bot deployed to Preview January 20, 2025 15:21 View deployment

Merge branch 'master' into feat/llamaindex-ts

3d33b70

vercel bot deployed to Preview February 3, 2025 08:55 View deployment

fix: replaced hardcoded acc id to dynamic in llamaindex.spec.ts

a08dfe4

entelligence-ai-pr-reviews bot reviewed Feb 3, 2025

View reviewed changes

vercel bot deployed to Preview February 3, 2025 09:52 View deployment

ellipsis-dev bot reviewed Feb 3, 2025

View reviewed changes

fix: add llamaindex/core

45efb71

vercel bot deployed to Preview February 3, 2025 10:16 View deployment

ellipsis-dev bot reviewed Feb 3, 2025

View reviewed changes

Merge branch 'master' into feat/llamaindex-ts

33fc664

vercel bot deployed to Preview February 11, 2025 06:56 View deployment

fix: smol_demo.py lint

fb89115

ellipsis-dev bot reviewed Feb 11, 2025

View reviewed changes

vercel bot deployed to Preview February 11, 2025 07:16 View deployment

Merge branch 'master' into feat/llamaindex-ts

f7e2760

vercel bot deployed to Preview February 11, 2025 07:17 View deployment

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat: add LlamaIndex TS support to JS SDK #1178

feat: add LlamaIndex TS support to JS SDK #1178

abhishekpatil4 commented Jan 10, 2025 •

edited by ellipsis-dev bot

Loading

vercel bot commented Jan 10, 2025 •

edited

Loading

entelligence-ai-pr-reviews bot Jan 10, 2025

entelligence-ai-pr-reviews bot Jan 10, 2025

entelligence-ai-pr-reviews bot Jan 10, 2025

github-actions bot commented Jan 10, 2025 •

edited

Loading

entelligence-ai-pr-reviews bot commented Jan 10, 2025

ellipsis-dev bot left a comment

shreysingla11 Jan 10, 2025

shreysingla11 Jan 10, 2025

shreysingla11 Jan 10, 2025

shreysingla11 commented Jan 10, 2025

plxity commented Jan 10, 2025

entelligence-ai-pr-reviews bot Jan 10, 2025

entelligence-ai-pr-reviews bot Jan 10, 2025

entelligence-ai-pr-reviews bot Jan 10, 2025

ellipsis-dev bot left a comment

ellipsis-dev bot Jan 10, 2025

entelligence-ai-pr-reviews bot Jan 10, 2025

entelligence-ai-pr-reviews bot Jan 10, 2025

ellipsis-dev bot commented Jan 20, 2025

entelligence-ai-pr-reviews bot Feb 3, 2025

ellipsis-dev bot left a comment

ellipsis-dev bot left a comment

ellipsis-dev bot left a comment

feat: add LlamaIndex TS support to JS SDK #1178

Are you sure you want to change the base?

feat: add LlamaIndex TS support to JS SDK #1178

Conversation

abhishekpatil4 commented Jan 10, 2025 • edited by ellipsis-dev bot Loading

vercel bot commented Jan 10, 2025 • edited Loading

entelligence-ai-pr-reviews bot Jan 10, 2025

Choose a reason for hiding this comment

entelligence-ai-pr-reviews bot Jan 10, 2025

Choose a reason for hiding this comment

entelligence-ai-pr-reviews bot Jan 10, 2025

Choose a reason for hiding this comment

github-actions bot commented Jan 10, 2025 • edited Loading

JS SDK Coverage Report

entelligence-ai-pr-reviews bot commented Jan 10, 2025

ellipsis-dev bot left a comment

Choose a reason for hiding this comment

shreysingla11 Jan 10, 2025

Choose a reason for hiding this comment

shreysingla11 Jan 10, 2025

Choose a reason for hiding this comment

shreysingla11 Jan 10, 2025

Choose a reason for hiding this comment

shreysingla11 commented Jan 10, 2025

Code Review Summary

Overall Assessment

Positive Aspects

Suggestions for Improvement

Code Quality Rating: 8/10

plxity commented Jan 10, 2025

entelligence-ai-pr-reviews bot Jan 10, 2025

Choose a reason for hiding this comment

entelligence-ai-pr-reviews bot Jan 10, 2025

Choose a reason for hiding this comment

entelligence-ai-pr-reviews bot Jan 10, 2025

Choose a reason for hiding this comment

ellipsis-dev bot left a comment

Choose a reason for hiding this comment

ellipsis-dev bot Jan 10, 2025

Choose a reason for hiding this comment

entelligence-ai-pr-reviews bot Jan 10, 2025

Choose a reason for hiding this comment

entelligence-ai-pr-reviews bot Jan 10, 2025

Choose a reason for hiding this comment

ellipsis-dev bot commented Jan 20, 2025

entelligence-ai-pr-reviews bot Feb 3, 2025

Choose a reason for hiding this comment

ellipsis-dev bot left a comment

Choose a reason for hiding this comment

ellipsis-dev bot left a comment

Choose a reason for hiding this comment

ellipsis-dev bot left a comment

Choose a reason for hiding this comment

abhishekpatil4 commented Jan 10, 2025 •

edited by ellipsis-dev bot

Loading

vercel bot commented Jan 10, 2025 •

edited

Loading

github-actions bot commented Jan 10, 2025 •

edited

Loading