Reright
Human approval for agent-written text
Reright makes your AI agents get your approval before they send text a person will read: commit messages, PR descriptions, emails, support replies. The agent submits a draft with full context (what it's about, who will read it, why it wrote it). You edit it into your own words, approve it, and the agent sends exactly what you approved. It works through an MCP server plus hooks for Claude Code, Codex, Gemini CLI, Copilot and Cursor, with a global git hook as backstop. Privacy-conscious by design: the hook only sends the SHA-256 hash of the text to check approval, never the text itself. The command-line client is open source (MIT). The hosted review service has a free plan (10 messages/month); paid tiers start at $5/month. Honest caveats: only Claude Code has been tested end-to-end; other agents' hooks are built from vendor documentation. The hooks run on your machine and can be bypassed by whoever can edit them. The server is not end-to-end encrypted: interpt operates it and could technically read your drafts.
Version:
Quick answer
Reright makes your AI agents get your approval before they send text a person will read — commit messages, PR descriptions, emails, support replies. The agent submits a draft with its context, you edit it into your own words and approve it, and the agent sends exactly what you approved.
At a glance
| What it is | A human-approval gate for agent-written text, built on an MCP server plus hooks, with a global git hook as backstop. The hook only sends the SHA-256 hash of a text to check approval, never the text itself. |
|---|---|
| Who it's for | People who let agents write outward-facing text and want a human checkpoint before any of it is sent. |
| What you give it | Your review: each draft arrives with full context (what it is about, who will read it, why the agent wrote it), and you edit and approve it — nothing sends before that. |
| What you get | An approval step in front of agent-sent text: drafts you can rewrite in your own words, and an agent that sends exactly the version you approved. |
| Works with | An MCP server plus hooks for Claude Code, Codex, Gemini CLI, Copilot and Cursor. The product notes that only Claude Code has been tested end-to-end; the other hooks are built from vendor documentation. |
| Cost | Freemium · free plan — the hosted review service's free plan covers 10 messages/month and paid tiers start at $5/month; the command-line client is open source (MIT). |
| Time | Setup of the MCP server and hooks once; after that, a review takes the time you spend reading and editing each draft before it is sent. |
Example — approving a support reply
An example of the documented approval flow:
- Your agent drafts a reply to a support email and submits it through Reright with its context: what the message is about, who will read it and why the agent wrote it that way.
- You read the draft, rewrite two sentences into your own words, and approve it.
- Only then does the agent send the reply — exactly the version you approved, checked by hash rather than by sending your text itself to the checker.
Install
Copy the install package below, then paste it into Muse🧪 Install ran in our test sandbox · Nov 11, 2026📌 Pinned version: commit 926e461Pinned version: install the code at commit 926e46118635c46b8408d74b7fbefca444247360 exactly (Skill Harbor tested this version). Do not substitute a newer version. If this commit is unreachable, stop and tell me. Get human approval on your agents' writing with Reright. 1. Create a free account at https://app.reright.it (no credit card needed). Note: the hosted review service's free plan is limited to 10 messages/month; paid tiers start at $5/month. 2. In the dashboard, copy your one-time setup block. 3. Paste it into your agent (Claude Code, Codex, Gemini CLI, Copilot or Cursor). The installer sets up the hooks, the MCP connection and the rules text. 4. Approve the test draft that appears in your browser. Done: your agent now submits its text before sending. Open-source CLI (MIT): https://github.com/interpt-co/reright-cli
Saved to your recent installs. Find it anytime on /connect.
🔒 Security, provenance and what this touches
The install prompt below already includes the vetting steps: your agent follows the community checklist before installing anything with executable code. Want more?
Verification certificate
Dated facts about what was checked for this listing, and how. This is not a safety certification.
- Pinned version
- 926e461 · 2026-10-13
- Sandbox install test
- Install ran in our test sandbox · Nov 11, 2026
- Creator identity
- Not verified
- Install prompt analysis
- 2026-10-13 : No commands, network calls, packages, file writes or API keys detected.
Skill Harbor never reviews the code of a listed skill. These are dated facts about the install prompt and our install test, not a safety certification. Review a skill yourself before you install it. How we verify
Questions
How do I install a skill?
Every product page includes a copy-paste install prompt. Paste it into your Muse and it sets the skill up for you — no manual configuration.
Where does my money go?
Straight to the seller. Skill Harbor never processes payments: checkout happens on the seller’s own page, usually Stripe.
What does the ✓ next to a creator’s name mean?
It means we confirmed the identity of the person behind the listing. It says nothing about the code itself — always check a skill before installing it.