Toolkit pick · harness
Agent Looper
Agent Looper is a harness in the 2026 agents.dancingteeth catalog. Goal-driven fix-until-green loops with a shell verifier — BYOK through Cursor, Cline, OpenCode, Pi, Codex, DSH, Muse, or Claude as the worker.
Last updated
Verified
Best for
Goal-driven fix-until-green loops with a shell verifier — BYOK through Cursor, Cline, OpenCode, Pi, Codex, DSH, Muse, or Claude as the worker.
Skip if
You want Cursor /loop or Automations (scheduled same-chat wakes), or you only need a one-shot agent turn.
Who it fits
- Goal-driven fix-until-green loops with a shell verifier — BYOK through Cursor, Cline, OpenCode, Pi, Codex, DSH, Muse, or Claude as the worker.
- Terminal-first engineers who live in the shell and git
Pros
Heard about agent loops? Fix-until-green with a frozen GOAL.md — fresh worker each iteration, shell verifier decides done, optional judge. Uses coding agents you already pay for (Cursor, Cline, OpenCode, Pi, Codex, DSH, Muse, and Claude); detects what is installed; you or your agent pick model and provider. OpenCode and Pi can route via OpenRouter, Vercel, Ollama, or another OpenAI-compatible router. Not Cursor /loop (that is a timer in the same chat).
Cons
Editorial Toolkit pick — not an Agents-shelf coding agent or Inference-shelf provider. See Find more for broader discovery.
Facts
- Locality:
- Cloud-native
- Surfaces:
- CLI
- Maturity:
- General availability
- BYOK:
- Bring your own API key
- Open source:
- Source available
- Verified:
- Sep 2026
Agent standards & memory
Compare one neighbor first
Vendor pages sell hard. Read one alternative on this site, then leave for the official link if the fit still holds.
Official links
Alternatives
Frequently asked questions
What is Agent Looper?
- Agent Looper is a harness covered in the agents.dancingteeth 2026 catalog. Heard about agent loops?
Who is Agent Looper best for?
- Goal-driven fix-until-green loops with a shell verifier — BYOK through Cursor, Cline, OpenCode, Pi, Codex, DSH, Muse, or Claude as the worker.
When should I skip Agent Looper?
- You want Cursor /loop or Automations (scheduled same-chat wakes), or you only need a one-shot agent turn.
What are the main trade-offs of Agent Looper?
- Pros: Heard about agent loops? Cons: Editorial Toolkit pick — not an Agents-shelf coding agent or Inference-shelf provider.
Does Agent Looper support AGENTS.md, SKILL.md, or MCP?
- This card lists AGENTS.md as N/A, SKILL.md as N/A, and rules/memory as GOAL.md + loop.json + log.ndjson; optional reviewGate judge.. Free options: MIT — pnpm add -D @dancingteeth/agent-looper @cursor/sdk.
Someone asks you about vibe-coding software? Share this site with them — we are adding more useful data regularly. agents.dancingteeth.net