AJX

AJX is an open-source skill and CLI for agent usability testing across digital products.

AJX (Agent Journey Experience) is an open-source skill and CLI that tests how agents use websites, command-line interfaces, software development kits, and application programming interfaces. It records agent attempts, tracks execution time, token usage, and retries, and checks the final result. Users provide a prompt, link, and success criteria, then review trials in a local web console. It supports testing with tools like jq, curl, SQLite, and Python Requests. The tool runs on macOS or Linux and provides JSON and JUnit results for continuous integration pipelines.

AJX at a glance

Best for: Software engineer, Product manager
Runs on: macOS, Linux, CLI
Pricing: Open source
Released: November 2026
Category: AI agent tool
Website: ajx.dev

Key features

Run usability tests for websites, command-line tools, and application programming interfaces
Record agent attempts, elapsed time, token usage, and retries
Open a local web console to review tasks, checks, and environments
Inspect timelines showing agent forks, blocks, waits, and recoveries
Generate product asks and fix recommendations based on event logs
Execute automated trials in continuous integration pipelines with ajx ci

Use it for

Test how an agent completes a task using a developer tool or command-line interface
Find friction points and barriers in website documentation for automated workers
Compare task execution results across different models and environments
Generate JSON and JUnit test reports for continuous integration workflows

Alternatives to AJX

AGiXT — Builds autonomous agents with integrated memory and web capabilities.
OpenViking — Unifies agent memory, knowledge RAG, and skills into a single directory structure.
Composio — Equips AI agents with tools, managed user auth, and sandboxes across 1,500+ apps
Skill-Up — Measures behavior across cases and compares runs with and without skills.
Adyen — Share AI agent resources across coding tools and repositories easily.
ARTEX — Manages autonomous penetration testing tasks and tracks vulnerability progress.
BrowserOS — Runs entirely locally and lets AI agents use real user logins to complete tasks.
iFixAi — Audits agent behavior independently without building manual eval harnesses.
MyAgent — It features built-in permission controls, long-term memory, and skill self-evolution.
Mindwalk — Visualize how coding agents explore and edit files on a 3D repository map.
Beeuni — Agents retain context and improve skills over time through a controlled review process.
ARTEX — Runs autonomous multi-agent penetration testing with a dual-graph architecture.

All AJX alternatives