mirror of
https://github.com/vxcontrol/pentagi.git
synced 2026-07-20 23:57:11 -04:00
[PR #17] [MERGED] Add ctester utility for testing LLM agents #122
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
📋 Pull Request Information
Original PR: https://github.com/vxcontrol/pentagi/pull/17
Author: @asdek
Created: 3/24/2025
Status: ✅ Merged
Merged: 3/24/2025
Merged by: @asdek
Base:
master← Head:14-enhancement-openrouter-configuration📝 Commits (1)
7da46effeat: added ctester utility for testing LLM agents, updated Dockerfile to build ctester and added provider configuration files, updated README.md with instructions for using ctester📊 Changes
17 files changed (+2978 additions, -37 deletions)
View changed files
📝
.vscode/launch.json(+18 -0)📝
Dockerfile(+10 -1)📝
README.md(+142 -0)➕
backend/cmd/ctester/main.go(+248 -0)➕
backend/cmd/ctester/models.go(+109 -0)➕
backend/cmd/ctester/report.go(+208 -0)➕
backend/cmd/ctester/tests.go(+833 -0)➕
backend/cmd/ctester/utils.go(+120 -0)📝
backend/go.mod(+1 -1)📝
backend/go.sum(+2 -2)📝
backend/pkg/providers/provider/wrapper.go(+48 -5)➕
backend/pkg/queue/queue.go(+209 -0)➕
backend/pkg/queue/queue_test.go(+228 -0)➕
examples/configs/deepinfra.provider.yml(+84 -0)📝
examples/configs/openrouter.provider.yml(+26 -28)➕
examples/tests/deepinfra-report.md(+346 -0)➕
examples/tests/openrouter-report.md(+346 -0)📄 Description
Description of the Change
This PR introduces the
ctesterutility for testing LLM agents. The utility enables developers to validate and optimize LLM agent performance through automated testing.Problem
The project lacked a dedicated utility for testing LLM agents, making it difficult to validate their functionality and optimize model selection.
Solution
ctesterutility with comprehensive test suites for various agent typesCloses #14
Type of Change
Areas Affected
Testing and Verification
Test Steps
ctesterwith various agent typesDocumentation Updates
Deployment Notes
No special considerations for deployment; the utility is built during Docker image creation.
Additional Notes
This utility will help improve LLM agent reliability by identifying optimal model configurations for different agent types.
🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.