Kojen

Full AI agent lifecycle platform — create, grade, optimize, deploy and orchestrate AI agents from your IDE or terminal.

Enables testing AI safety classifier robustness against query decomposition, obfuscation, and multi-agent attacks. Provides tools for full evaluation pipelines, query previews, and status checks.

Let coding agents play, test and tune your 2D web game. The debuff MCP server attaches to the game you have open (or runs a headless copy) and gives agents live state, explained and source-mapped errors, frame timings, typed params, deterministic replays and rewind, scenario tests and fuzz playtesting. Everything an agent does shows in your dev panel, where you can undo it.

Enables red-team testing of AI agents against indirect prompt injection by fuzzing payloads with delimiter evasion and Unicode homoglyphs, detecting exfiltration canaries, enforcing least-privilege tool boundaries, and producing cryptographic audit proofs.