Every agent Agentastic supports

Agentastic's own terminal agent, with resume, unattended runs, a non-interactive exec mode, and built-in code review.

Prime Agent

Prime Intellect

Prime Intellect's open-source RLM agent, with persistent sessions, recursive subagents, a continual harness, and native ACP chat.

Anthropic's coding agent, launched in an isolated worktree with its native settings, skills, hooks, and MCP servers intact.

OpenAI's terminal coding agent, with resume, plan, unattended, and code-review workflows built into Agentastic.

Warp's terminal agent, launched as a TUI with unattended approval and its opening prompt typed once the interface settles.

Poolside's ACP-compatible coding agent, with queued interactive prompts, plan mode, session resume, and unattended execution.

A provider-agnostic terminal agent with sandbox, resume, unattended, and non-interactive review modes.

A dedicated review agent that analyzes the current worktree from Agentastic's code-review surface.

A repository-aware review agent available directly from Agentastic's code-review workflow.

Langbase's taste-aware coding agent, with its native configuration and skills available in every worktree.

Cursor's CLI agent, launched without Cursor's own worktree setup so Agentastic remains the isolation layer.

Google's terminal coding agent, with interactive prompts, plan mode, resume, and unattended runs.

Google's terminal-first Antigravity client, using your existing settings across isolated worktrees.

xAI's coding agent with prompt-file support for cloud machines and full interactive control on your Mac.

Meta's terminal coding agent, powered by Muse Spark 1.2, with seeded prompts, session resume, and read-only planning runs.

Alibaba's terminal coding agent, with plan, unattended, resume, and seeded interactive sessions.

Factory's software-development agent, available in normal or high-autonomy mode with session resume.

Sourcegraph's agentic coding tool, with stdin prompt delivery and direct access to the latest thread.

The open-source, provider-flexible coding agent, running with your models and configuration in each worktree.

GitHub's Copilot CLI, seeded into interactive mode and isolated from every other running agent.

Charmbracelet's terminal coding agent, with automatic prompt typing into its native TUI.

Augment Code's context-engine agent, launched with indexing enabled and support for continued sessions.

Block's open-source, extensible coding agent, with local and prompt-file delivery for remote machines.

Moonshot AI's terminal coding agent, supporting plan mode, unattended work, and directory-aware resume.

Kilo Code's terminal agent, available as another provider choice for isolated, parallel tasks.

Cognition's coding agent CLI, launched directly from Agent Home alongside local terminal agents.

AWS's spec-driven coding CLI, with trusted-tool execution and chat-session resume.

Atlassian's development agent, run through the Atlassian CLI with automatic prompt delivery.

Cline's terminal agent with plan mode and configurable automatic approval for isolated tasks.

Continue's open-source CLI, with read-only planning, session resume, and automatic runs.

Codebuff's terminal coding agent, launched with a task prompt and support for continued sessions.

The standalone Codebuff-compatible agent, with its initial prompt typed into the TUI and session continuation.

Mistral's open-source terminal assistant, with native unattended, plan, model-selection, and continued-session modes.

A compact, extensible coding-agent harness that keeps its native terminal interface inside Agentastic.

The established open-source pair-programming agent, with its interactive prompt typed in after launch.

Nous Research's open-source terminal agent, launched in its interactive TUI with resume and unattended execution.

Autohand AI's terminal agent, with unrestricted mode for intentionally hands-off workflows.

A memory-first coding agent that Agentastic starts as a fresh TUI conversation, with automatic prompt delivery.

The open-source software-development agent, with task prompts, automatic approval, and last-session resume.

Snowflake's terminal coding agent for development and data-engineering work.

Google's asynchronous coding agent CLI, with automatic task entry into its interactive interface.

JetBrains' model-agnostic coding agent, with Brave Mode, prompt seeding, and session resume.

The deliberately small SWE-bench agent, useful for focused tasks and reproducible experiments.

OpenBlock Labs' autonomous coding agent, with plan mode and latest-session resume.

MiMo Code

Xiaomi

Xiaomi's terminal coding agent, with seeded interactive prompts, unattended execution, and latest-session continuation.

Qoder CLI

Qoder

Qoder's terminal coding agent, with interactive prompt seeding, unattended execution, native plan mode, and session continuation.

OpenClaude

OpenClaude

An open terminal coding agent with positional prompts, unattended execution, native plan mode, and session continuation.

Ante

Ante

Ante's interactive coding TUI, with automatic prompt typing and a native unattended mode.

Oh My Pi

Oh My Pi

A terminal coding agent with positional prompts, automatic approval, model and thinking controls, and session continuation.

OpenClaw

OpenClaw

OpenClaw's self-contained local TUI, launched without a separate gateway and seeded through its message flag.

Tencent's terminal coding agent, with positional prompts, unattended execution, and session continuation.

A minimal terminal coding agent, launched with unattended approval and seeded once its interface settles.