Claude Code Subagent and Multi-Agent Mods

Pages your phone (ntfy or Telegram) when the coding agent finishes a long turn or needs you, and lets you reply from the phone

Радар фоновых агентов и живой канвас планов: что делает каждый агент, кто завис, кто ждёт тебя; общий план сессии с вложенными планами сабагентов…

Picks the model and effort of every subagent from its task type; with hot mode on, raises them when the 5-hour window resets with usage to spare

Shows the state of Claude (working, permission, input, idle, background, compacting) in the iTerm2 tab subtitle and in a Sessions pane in the desktop…

Alvearium: one ticket board for all your coding-agent sessions (Claude Code in the terminal and on claude.ai/code, and OpenAI Codex CLI) and their…

Claude Code subagents on the models you choose: name as many agents as you like, each on a model from your own providers (LM Studio, Unsloth Studio…

A Claude Code pane showing the model, thinking level, context, usage limits, git state, PR, running agents and your last three prompts

Live group chat for agents: several agents discuss in parallel in one shared room, with instant alerts, a silence timer that forces status updates…

Asks a local Clef decision model whether an expensive tool call or subagent is actually needed before Claude Code pays for it

Picks the Claude model and reasoning effort for each turn with Cloudflare Clef, a fast decision model, on Workers AI or fully local (llama-server…

Command Center: a live pane with context and plan limits, background work, subagent details, side questions, approvals and quick actions

Maps the conversation into branches: a pane of branch cards with Haiku titles and summaries, a live browser map with click-to-switch, and subagents…

Right-side Agents pane by Triple Digit (triple-digit.com, tailored AI solutions for each company): every working and recently finished session as a…

An always-on team of AI agents on your own computer: chat with them in the Crew window, and they work with your files, the web, your apps and the…

Deleting outside the project folder requires the user's approval (Bash, PowerShell, cmd, subagents, MCP)

Read-only pane that shows what Figma design work costs in a session: spend, screens, iterations per screen, failures and tokens by agent

Continuous engineering loop for Claude Code: Jev-based size/clarity triage, a plan-clarify-build-test-revise harness for small tasks, and for large…

Cost-free mode for Claude Code

A Claude Code mod that puts a live status band above the prompt: progress steps, context and rate-limit usage, subagents, a sun-arc clock with an…

A pixel-art HUD for Claude Code: context, cost, cache, agents and rate limits at a glance, little 8-bit scenes while Claude thinks, and Telegram…

Sets the model of every subagent spawn (agent.spawn)

Jev Semaphore for Claude Code: Jev / Kev model routing for subagents, with a live allocation pane, agent graph and context-exchange band

One orchestrator over every Claude Code session you have open, with Jev (TypeSafe System One) deciding which one needs you

Routes each subagent to a Claude model and effort that fits its task, judged by Jev

Routes each turn and subagent to a model tier and reasoning effort using a classifier behind any server that speaks POST /v1/classify…

Cheaper Claude Code sessions: read and edit by stable line IDs, a model router that sends each task and agent to the cheapest model trusted with it…

Raily is a personal AI agent that finds people worth meeting for you (friends, dates, business contacts) and handles the first contact

Selects a model and an effort level for Claude Code subagents

A /scopus-mod side pane for browsing the scopus-for-dobby library and attaching papers to the chat, with tools the agent can use to steer the same…

Sidebar with a running summary of the session: tokens, tools, skills, connectors, subagents and activity

Manage AI agent skills (SKILL.md), playbooks and value chains: list, toggle, categorise, install from GitHub/monorepo/ZIP/local/marketplace, update…

用量透视：5 小时窗口或任意时段内，各 session 的用量占比，及每个 session 里主 agent / 子 agent 各用了多少、什么模型

Keygrip work orders in a pane: /workorders lists your project's deploy queue; pick one to read its fields, edits, schema and verify steps, see the…

A pane, opened with /catalog, that lists every skill, command, agent, mod, plugin, settings hook, MCP server and tool by category: descriptions…

手機聊天室的授權卡 mod：桌面對話的 Bash 授權與選擇題同時送到手機，誰先按算誰；取代 settings.json 的 perm-bridge.js（mod 出錯時引擎自動退回它）

A dock above the prompt: plan usage windows, main and subagent context fill, and the shells running now with what each one is doing

convo-tree with real forks: switching a branch swaps the live transcript to that fork, so Claude sees only that branch from then on

Beobachtet nach jedem git push den Cloudflare-Build (Workers Builds), zeigt den Stand unter dem Prompt, gibt Claude bei einem roten Build das Log und…

Picks the model and the reasoning effort for each task with TypeSafe's Jev, a System One decision model reached either through TypeSafe's own API or…

Keeps personal data out of Claude: every tool output is scanned on your machine (OpenAI Privacy Filter and PII-Tracer), and output with personal data…

Spots research prompts and keeps them out of the main context: sends them to your own research endpoint (e.g

A workspace pane for Claude Code: agents and worktrees, checkpoints, notes, changes and memory files in one tabbed pane

サブエージェントと Workflow の agent() の依頼を LocalJev でロールに分類し、agent-routing の roles.json に従って model と effort を切り替える function hook

Spin up Claude Code workers from a picker: OpenRouter (any model) or Amazon Bedrock (Claude), as tmux panes or terminal tabs; team messaging between…

The main loop only delegates; the mod decides each subagent's model and effort from its brief, with Apple's on-device model and, when a key is set…

Choose the model versions Claude and its subagents may use: /models lists the newest of each family from Anthropic's Models API (older versions one…

Runs Explore subagents on haiku and general-purpose ones on sonnet unless a model was asked for

用 System One 模型（目前 TypeSafe Jev）做判斷的 hook：派 sub-agent 前選 model；grilling 收尾拆題成 context/specs 規格檔（需 TYPESAFE_API_KEY，停用 plugin 即關閉）