Disclosure: this article contains affiliate links. If you buy through them we may earn a commission at no additional cost to you. Our recommendations are based on the evidence recorded alongside this article, and prices were verified on the date shown.

A developer using OpenCode has to approve or deny a permission prompt every time the AI agent wants to run a command, edit a file, or touch the network. On a busy session that can mean dozens of interruptions. On 2026-08-16, a GitHub user going by huey77 published a plugin called opencode-auto-permissions and posted it to Show HN, aiming directly at that friction point.

What changed and why it matters

The plugin's stated job is narrow: automatically review OpenCode's permission prompts so you stop clicking "allow" on the same category of action over and over. huey77 describes the intended behavior as comparable to Claude Code's auto mode for permission handling, which gives OpenCode users a rough mental model of what to expect — an agent that filters low-risk, repetitive prompts instead of surfacing every one to a human.

Why it matters now: permission-prompt fatigue is one of the more common complaints from people running agentic coding tools for extended sessions. A plugin that specifically targets that friction, built and shipped as open source rather than bundled into a paid product, is worth knowing about before you build your own workaround or accept the default prompt volume as unavoidable.

Hands-on findings

There's no independent usage log to report here — this review is based on what the project's GitHub page and its Show HN posting state about the plugin's purpose and design, not on a run log or benchmark from using it ourselves. What the source establishes: the plugin sits inside the OpenCode plugin ecosystem, it intercepts permission prompts, and its comparison point is Claude Code's auto mode, which OpenCode users may already be familiar with from cross-tool experience.

What the source does not establish: no accuracy figures, no false-approval rate, no list of which permission categories it handles versus skips, and no version history are part of the cited material. If you're evaluating this for a team workflow, treat those as open questions to check directly in the repository's README and issue tracker before adopting it, rather than assuming parity with Claude Code's auto mode beyond the stated intent.

If your reason for reading this is actually about picking infrastructure to deploy the projects your agent builds — rather than the permission-prompt plugin itself — that's a separate decision, and it's where a hosting platform's tier structure comes in. [Check current price for Vercel Enterprise](Check current price for Vercel Enterprise) if your team's deployment needs go beyond what a self-serve plan covers.

Pricing and plans

opencode-auto-permissions itself carries no listed price in the cited source — it's distributed as an open GitHub repository, which is the standard distribution model for a free, self-installed plugin. There's no evidence of a paid tier, license fee, or usage cap attached to the plugin, so budget isn't a factor in adopting it; the cost is your own time spent reviewing and configuring it.

The comparison worth making explicit is between "free tool you self-host and self-manage" versus "platform you pay for and outsource the operational burden to." That's the actual decision most readers researching an OpenCode-adjacent tool are also weighing when it comes to where their code runs:

Exact dollar figures and per-tier limits for Vercel Pro and Enterprise aren't reproduced here because pricing pages change and the only reliable number is the one on the page at the moment you're deciding. [Check current price for Vercel Pro](Check current price for Vercel Pro) to see what's included at that tier before you commit.

Who it suits and who it does not

Best for: OpenCode users who run long agentic sessions and find themselves approving the same category of permission prompt repeatedly. If you already trust an LLM to make routine tool-use decisions — the same trust model that makes Claude Code's auto mode useful to its users — the plugin's stated approach should feel familiar rather than risky.

Skip it if: you work in a regulated or high-stakes codebase where every file write and command execution needs a human in the loop as a matter of policy, not preference. An auto-reviewing permission layer is, by definition, removing some of that human checkpoint, and the cited source doesn't document an audit log or override mechanism you could point to for compliance purposes.

Best for: teams evaluating whether to consolidate their deployment pipeline onto a single platform once their OpenCode-driven projects are ready to ship, and who want tier options that scale from a solo project to an organization.

Skip it if: you're already committed to a different hosting provider and switching costs outweigh whatever Vercel's current tiers offer — that's a call only the current pricing page, not this review, can settle for your specific workload.

Verdict

opencode-auto-permissions is a free, open-source answer to a specific, well-understood annoyance — permission-prompt fatigue in OpenCode — built by an independent developer and positioned explicitly against Claude Code's auto mode as a reference point. There's no cost barrier to trying it and no hands-on data here to tell you how well it performs beyond its stated intent, so the responsible move is to read the repository directly before wiring it into a workflow you depend on. If the reason you're here is really about where to deploy what OpenCode helps you build, [see current pricing at the merchant](See current pricing at the merchant) and match the tier to your team's actual deployment volume rather than guessing.