Top 200 ReposEXPLORER
Browse agents, categories and pages
Back to results
Reported by sourceThe source says this happened. Top 200 Repos has not reproduced it.

Rebuild a side project in OpenCode to test provider lock-in

Matthew Brunelle rebuilt a YouTube Music to OpenSubsonic shim with OpenCode and Kimi K2.6 to compare it with an earlier Claude Code and Opus 4.6 attempt.

Try the promptRead the original (opens in a new tab)

The problem

How it works

He repeated an earlier experiment using OpenCode with Kimi K2.6 through OpenCode's Go subscription, to see whether an open harness and open-weight model could stand in for Claude Code. He used the same spec-then-review approach, pasted server logs back to fix client-specific issues, and noted differences in planning, permissions and tool use.

Results and limitations

He says the code came out roughly as good as with Claude Code, and OpenCode tested the server with curl unprompted, but it lacked a real plan mode, ran tools without asking by default, rarely used its question tool, and hit several network-lost errors. He still felt more in control with Claude Code, yet cancelled his Anthropic subscription for now.

Before you try it

One project and one model pairing; he cannot tell whether some tool-use issues came from Kimi K2.6 rather than OpenCode. He shelved self-hosting because the large open models need far more memory than he has.

Tools and repositories

Tools
OpenCode · Kimi K2.6 · yt-dlp · Claude Code
Prerequisites
Needs a model provider account; OpenCode's tools run without approval by default, so tighten permissions first.

Try it yourself: Run a side-by-side coding agent bake-off

Fill the fields to adapt the prompt. Empty fields stay as visible [placeholders].

Preview
I want to compare coding agents on the same task. Read [Spec file] for a small project: [Project summary]. Work on a fresh branch named [Branch name]. Before coding, write a short plan file and wait for my approval. Ask me before assuming anything about authentication, data reuse or external services. Implement in small steps, run [How to test] after each, and keep a log in [Agent log file] of every point where you needed my help, retried a tool, or changed the plan. Do not push, publish or install global packages. When finished, review the code against the spec with fresh eyes and list any gaps.

Give every agent the same spec and permission settings so differences reflect the tool, not the setup.

Related videos

About anomalyco/opencode

About anomalyco/opencode

Independent creators. Listing a video implies no endorsement, sponsorship or partnership.

Sources

  1. Trying Out OpenCode To Avoid Claude Code Lock-in (opens in a new tab)

    First-person account · blog.matthewbrunelle.com · Published 30 Apr 2026 · Source checked 9 Nov 2026, 22:23 UTC

    What this source supports
    • Used OpenCode with Kimi K2.6 to rebuild an OpenSubsonic shim
    • Compared with an earlier Claude Code and Opus 4.6 run
    • Code quality roughly on par; missing plan mode and persisted permissions
    • Network-lost tool errors; cancelled Anthropic subscription for now
Report a problem

Reported by source

Explore and change a large codebase from the terminal with Gemini CLI

Gemini CLI is Google’s open-source terminal agent. Simon Willison found it strong on large codebases in 2025; in 2026 Google moved individual users to Antigravity CLI.

github.comgoogle-gemini/gemini-cli
Reported by source

Build an app with Replit Agent while keeping production data out of reach

Replit Agent builds and browser-tests apps from a description. A widely reported 2025 incident, where it deleted a live database during a code freeze, shows why separate dev and prod data matters.

docs.replit.com
Reported by source

Hand coding tasks to Cursor cloud agents that open pull requests

Start Cursor agents from the editor, web, Slack or a GitHub comment; each works in its own cloud VM on a branch and returns a pull request with screenshots, videos and logs.

cursor.com
Planned, not done

Run coding agents from one self-hosted control centre with OpenHands

OpenHands Agent Canvas runs the OpenHands agent, or Claude Code, Codex, Gemini CLI and other ACP agents, on your machine or in Docker sandboxes, and can trigger them from schedules or webhooks.

github.comOpenHands/OpenHands
Reported by source

Ship a non-trivial app feature over many short agent sessions

Mitchell Hashimoto built Ghostty's unobtrusive macOS update notifications with the Amp agent over 16 sessions, rewriting and reviewing parts by hand before merging.

mitchellh.com · 11 Oct 2025
Reported by source

Apply the context engineering lessons from Manus to a long-running agent loop

The Manus team describes the context rules it settled on after rebuilding its agent framework four times: keep the prompt prefix cacheable, mask tools instead of removing them, and use files as memory.

manus.im · 18 Jul 2025