Overview
OpenCode Permission Reviewer is an unofficial community plugin for OpenCode. It provides a tool-free AI reviewer for every ask permission. The reviewer reads the request, the user's policy, and the session context, then allows, denies with feedback, or escalates to the user. This aims to let safe actions proceed without waiting for a keystroke while blocking or surfacing genuinely risky ones.
The plugin adds a Codex-Guardian-style reviewer: a dedicated, tool-free model backend reads the pending request, bounded transcript evidence, recovered user intent, and a tenant policy. Actions classified as critical are not auto-approved. Failures do not become new approvals; they lead to manual review or denial.
It preserves policy: allow continues, deny stays blocked. Tool-free review means normal models run in an isolated scratch session with every tool denied; Jev uses its typed System One API. Read-only enrichment provides bounded, sanitized SSH/local-script/Git evidence. It is auditable with one JSONL record per review, remote commands stored as SHA-256. An optional TUI overlay shows review state.
Pricing
Free
Detailed plans have not been confirmed in our catalog. Check the official website for current limits and billing terms.
Visit WebsitePrices and limits may change. Confirm the currency, billing period, seat minimum and usage caps on the official website.
Use Cases
- Teams use OpenCode Permission Reviewer to automatically approve safe routine actions without waiting for a keystroke.
- Teams use it to block or escalate genuinely risky actions.
- Teams use it to enforce a tenant policy for permissions.
- Teams use it to get auditable records of permission reviews.
- Teams use it to review permissions with read-only enrichment from SSH, local scripts, and Git.
Features
AI assistant
Natural language interface
Workflow automation
Reviews
No reviews yet. Be the first to share your experience!