agent-supervision

Temporary rooms for Humans and AI Agents — run, supervise and steer Agent Tasks without a permanent workspace.

An agent supervisor and software factory foreman powered by decision models

Open-source agent execution-recovery controller for coding and tool-using AI agents.

ChatGPT <-> Codex bridge: local coding agents ask ChatGPT for advice, and ChatGPT can inspect or steer your running Codex sessions. No API key.

Farming Code is a self-hosted browser workspace for running and supervising Codex, Claude Code, OpenCode, and other AI coding agents.

Open-source controller and tamper-evident evidence layer for long-running AI workers

Human-in-the-loop control surfaces for monitoring and directing AI agents across Novation Launchpads, browsers, phones, and tablets.

Delegate coding tasks to agents, collect artifacts, and verify results locally with standalone supervision skills for Claude, Codex, and Kimi automation.

Send Pi agents back to work when they stop before the job is done.

Windows-first local agent supervision skills for Claude Code, Codex, and Kimi Code. 本地 Agent 委托审查与可追溯监督。

AI coding that actually makes you move with superhuman speed.

DSH DevKit: multimodal gallery + multi-agent supervision console (DeepSeek Harness dynamic Cordis plugin)

MCP server for Cursor Cloud Agents (background agents): launch, follow up, supervise and track the cost of Cursor agents from Claude Code, Codex CLI or OpenCode.

User-intent tracking and agent supervision for DeepSeek Harness (DSH) · 用户意图监督与受控纠偏

Supervise an AI coding agent (Claude Code / Codex / ACP) overnight and keep control: block forbidden tool calls before they run, gate 'done' on real tests, checkpoint work to git, cap spend, and get a morning report where every line is sourced. Local-first, no telemetry. Rust daemon + Tauri app + Claude Code plugin.

DSH task supervision with DAG progress, plan review, persistent control and stage/completion checks.

The control tower for your coding agents — one resident mind above Claude Code, Codex, opencode, zcode: every session, the machine's vitals, and the disk's actual record in one attention state. Verdicts SILENT·FLAG·ASK·DENY — never allow. Watch→contest→advise→drive. C/C++20, local spine, hash-chained ledger, 9B on your RTX. FUSOR-1. MIT.

Independent supervision and evidence gates for long-running coding-agent work

Calibrated yes-or-no checks for AI coding agents: a Claude Code plugin and a Rust crate on the Jev decision model

Autonomous CTF batch solver with 3-layer LLM supervision — Claude Code in Docker per challenge, deterministic watchdog + flag gate, operator babysit playbook