StableLib benchmarking suite
A tiny guard that forces tests and benchmarks to run inside Docker.
Let Claude Code, Codex and Cursor use your signed-in browser, several agents at once, each in its own tabs. Fastest of four tools in a published benchmark.
A Karma benchmark reporter
Let Claude Code, Codex and Cursor use your signed-in browser, several agents at once, each in its own tabs. Fastest of four tools in a published benchmark.
Benchmark your JS projects with nanosecond resolution
Talk to Excel in DeepSeek Harness: read, edit, clean and report on spreadsheets by conversation. Re-checks every formula after each edit, repairs silent errors, and needs no local Excel for the core work (charts and PDF export use Windows Excel).
Lightweight reporting SDK for evo experiments. Zero dependencies.
Let Claude Code, Codex and Cursor use your signed-in browser, several agents at once, each in its own tabs. Fastest of four tools in a published benchmark.
Can a language model predict how a crowd reads a word? A cultural fluency benchmark built from Mixed Messages.
Evidence-led web anti-cloning toolkit: server-side quota/pagination guards, Web Bot Auth (RFC 9421) signed-agent verification, canary traps, a Python/FastAPI quota-guard port, controlled twin-SaaS benchmark and REA envelope guidance.
Performance benchmarking tool for headless browser automation. Measure CPU, GPU, timing, and screenshot performance across configurations.
Let Claude Code, Codex and Cursor use your signed-in browser, several agents at once, each in its own tabs. Fastest of four tools in a published benchmark.
- Fastest JSON deep clone implementation for Node.js - Goal of this project is to track and maintain the fastest deep cloning function for plain JavaScript objects (POJOs), aka JSON objects. - Keeps track of benchmarks of all other implementations.
Benchmark page: measure what one node can carry (the orchestrator's node benchmark), inspect one image or one sound
Two functions that measure the computation power
A simple benchmark for Storybook. Usage:
Let Claude Code, Codex and Cursor use your signed-in browser, several agents at once, each in its own tabs. Fastest of four tools in a published benchmark.
Cli for running benchmark tests for n8n
Advanced AI coordination framework with MCP protocol integration
Run performance benchmark tests
A simple suite of benchmarking tests
The Multica connector for 10xagents: sync a setup's agents, squad and workflow to a Multica workspace, and benchmark it.
LDF Server SPARQL endpoint hook handler for JBR
Production-ready AI agent library using AI SDK v6 ToolLoopAgent for GAIA benchmarks with swappable providers
Benchmark agent configurations (harness + model + effort + profile) on your real workflows.
Benchmark.js results in ascii tables for NodeJS
Verdict evaluation engine: rubrics, ingestion, judges (LLM / human / benchmark), aggregation.
Usefull functions for solving programming tasks
OpenCode plugin for skill creation — custom tools for validation, evaluation, description optimization, benchmarking, and review serving.
Personal tools to post-process query results from SPARQL benchmarks
Benchmark a DSH agent harness, not a model: run one fixed task against controlled pipeline variants (tools, persona, injected tool faults, model route) and record deterministic per-trial traces as JSONL.
Core contracts for @ac-bench: condition and case registration, the per-condition runner, the MeasureData wire shape, and the measure and reporter plugin interfaces.
Protocol verification — RFC compliance tests, browser golden packet captures, integration tests, and benchmarking.
javascript standard data structure library which benchmark against C++ STL
Benchmarking library, based on Benchmark.js and Autocannon
Group and sort Eleventy’s verbose output by directory (and show file size with benchmarks)
This package feeds metrics, marks and measures collected by the `react-native-performance` library to the `Performance` Flipper plugin.
Display task stats summary for Gulp
Host-agnostic Winston logger with lazy logging, benchmarks and a small host interface. Unofficial; not affiliated with Winston.
Ensures that new code does not introduce performance regressions with CI. Tracks: