Find what is wrong with the AI agent configuration your repository already has

Find what is wrong with the AI agent configuration your repository already has — doctor, check, lint, audit, adopt, compile, eval

Read, analyse and compile AI agent configuration — discovery, a normalized IR, a resolver, diagnostics, and static security analysis

Local dashboard and API for agentfile (beta)

Iris agent image toolchain — Agentfile model + parsers, builder (resolve/embed/pin/validate), content-addressed inspectable image + OCI layout, lockfile, integrity verify, and session pinning + definition migration. Host-side, zero external deps.

CLI for defining and running LLM workflows as files — for any IDE agent.

Bourdon credential-redaction SSOT + federation leak auditor — the security keystone of the TS mirror. 12 keyword + 14 token credential patterns (exact per-pattern case flags), redactText/containsSecret, and the static L5 leak audit (credential + visibilit

AI coding agent config lifecycle manager — keeps your agents from forgetting your project

Runtime for defining and executing portable AI workers. Multi-provider LLM support (Anthropic, OpenAI, Gemini, and OpenAI-compatible vendors), pluggable persistence and tracing adapters, an MCP tool client, streaming, abort, and pause/resume.

React renderer for the agent file format (.agent).

Den CLI — the living runtime for AI agents. Connects to local Docker agents and remote deployed agents (Northflank, Cloudflare, etc.) from one terminal. Pairs with the den-agent Python runtime on PyPI.

Library to automatically symlink AGENTS.md based on configuration

Skill manager and sync engine for AI coding tools — sync skills, MCP servers, rules, and commands across 50+ agents

Self-contained HTML agent format with built-in security

Webpack plugin that injects quello, the visual element picker for AI coding agents.

Vendor-neutral, replay-safe model adapter — the OpenAI model_call performer behind Iris's portable model port; a peer of @irisrun/provider-anthropic behind the same port, so swapping providers leaves the agent and its byte-identical journal replay unchang

Vendor-neutral, replay-safe model adapter — the Anthropic model_call performer behind Iris's portable model port; swap providers without touching the agent, and the captured reply replays byte-identically from the journal. Direct Anthropic Messages over b

CLI to manage AI agent configuration files across Claude and Cursor platforms

Canonical Cedar schema library for agent action verbs (exec, open, connect, request_tool). TypeScript helpers for constructing Cedar requests against the schema.