Orchestration

Delegate coding tasks to Codex, Claude Code or OpenCode as background workers

Dispatch two or more independent tasks to run in parallel agents

Run implementation plans as independent tasks handed to subagents

Pursue a long goal in one session by coordinating several subagents

Plan agent-run engineering with eval-first checks and cost-aware model routing

Build a persistent multi-agent setup with file memory, commands and schedules

Run Claude Code as an autonomous agent with memory, schedules and a task queue

Review patterns for autonomous loops, from sequential pipelines to multi-agent DAGs

Dispatch parallel coding agents into isolated worktrees and track their reports

Run autonomous agent loops that self-check, gate on evals and recover from failures

Get four distinct voices to argue an ambiguous decision before you choose

Add a critique from a separate model to stress-test a draft decision

Simulate a dev team of PM, architect, developer and QA personas on one problem

Run parallel agent sessions in tmux panes across Claude Code, Codex and other harnesses

Convene a four-voice council to weigh tradeoffs before a go or no-go call

Build features through generator and evaluator agent loops until they clear a bar

Check an agent loop design for spinning, verifier gaming and runaway runs

Split a task into parallel lanes with isolated writes and a final verification check

Split an RFC into dependent work units and drive each from plan to tested merge

Run repeated rollouts and gate promotions through an append-only decision ledger

Gate outputs behind two independent reviewers who must both pass before release

Coordinate agent squads with owners, branch isolation and merge gates in parallel

Pick up to five agent personas and run them in parallel on one task

Settle merge conflicts between two coding agents with a neutral arbiter

Hand coding tasks off to the Blackbox AI multi-model command line tool

Delegate feature work and pull requests to the Claude Code CLI

Delegate feature work and pull requests to the OpenAI Codex CLI

Plan parallel agent fan-outs in code with adversarial verification waves

Delegate feature work and pull requests to the xAI Grok Build CLI

Delegate coding features and PR reviews to the OpenCode CLI

Delegate coding tasks to the OpenHands CLI with any model through LiteLLM

Handle planned task follow-ups, background task completions and task creation rules

Create subagents with frontmatter, system prompts and triggering descriptions

Split large tasks across parallel workers with a stronger model reviewing the plan

Run the OpenAI Codex CLI in three modes from inside your coding session

Delegate code location, small edits and diff reviews to compressed-output subagents

Coordinate supervised agent workers with task graphs, decision gates and handoffs

Autonomously submit brief, raw complaints about agent tooling without asking first

Delegate requested work to an existing Paperclip agent with a durable task reference

Find progress or deliverables from earlier Paperclip work and add feedback

Read task context, deliver files, update blockers and run Paperclip control plane actions

Manage a Paperclip company as a board member through chat, including approvals

Draft and submit hire requests for new Paperclip agents with governance checks

Summarize a Paperclip team's agents, tasks, blockers and pending approvals

Create Paperclip status cards and turn interest prompts into bounded search queries

Quietly submit concrete improvement ideas for repeated agent friction without asking

Route free-form requests to the right task type and ask only when it is unclear

Plan requests and run builds with a bundled strategy entry point for an agent

Commit an agent to a written task plan before it starts generating output