GitHub launch kit

April 5, 2026 · View on GitHub

This file is the source-of-truth copy deck for the GitHub-facing launch story for VeriClaw 爪印.

Use it when writing:

  • the GitHub release body
  • the repository README launch section
  • short launch blurbs
  • FAQ replies
  • screenshot captions and launch-safe positioning
  • release hero art copy
  • pinned repo blurbs
  • first-wave distribution posts

Current release posture:

  • GitHub is the current public launch surface
  • Apple companion hardening continues in parallel
  • do not imply the App Store path is already publicly live

Positioning

VeriClaw 爪印 is the Apple-native correction companion for OpenClaw.

OpenClaw remains the runtime, gateway, and assistant substrate. VeriClaw adds the native supervision layer:

  • evidence-first correction cases
  • professional-role drift diagnosis
  • prescription and follow-up guidance
  • verification before closure
  • per-bot casebook learning
  • a desktop hover companion that keeps supervision pressure visible

Short copy

Hero line:

  • Apple-native correction companion for OpenClaw

Short subtitle:

  • Turn drift evidence into diagnosis, intervention, verification, and casebook learning.

One-sentence description:

  • VeriClaw 爪印 is the Apple-native correction companion for OpenClaw, designed to supervise drifting bots through evidence, diagnosis, prescription, verification, and casebook learning.

Medium description:

  • VeriClaw 爪印 strengthens OpenClaw with a native correction workspace. Instead of stopping at alerts, it helps turn evidence of hallucination, fake completion, disobedience, or role drift into a concrete loop: diagnosis, recommended intervention, verification, and a per-bot casebook that improves future corrections.

Long description:

  • VeriClaw 爪印 is an Apple-native correction workspace and supervision companion for OpenClaw. It is built for the moment when a bot is drifting, overreaching, hallucinating, or failing its professional role contract. Rather than only showing telemetry, it organizes the problem into evidence, diagnosis, prescription, verification, and casebook updates. The result is a companion surface that helps teams understand what went wrong, what to do next, and whether the loop is actually closing.

Launch pillars

Keep these four ideas together in most public GitHub-facing copy:

  1. Companion, not clone
    • VeriClaw strengthens OpenClaw instead of pretending to replace it.
  2. Correction, not just monitoring
    • the product should answer what to do next, not only what went wrong.
  3. Professional-role drift
    • the app treats many failures as occupational-role contract failures when appropriate.
  4. Apple-native supervision
    • the macOS app, iPhone companion, and hover surfaces should feel like one native workflow.

GitHub-first framing

When the public message needs to reflect the current rollout order, prefer:

  • GitHub-first launch for the OpenClaw + VeriClaw story
  • public repository launch now, Apple submission path continuing in parallel
  • GitHub is live first; the Apple companion follows after final hardening

Avoid:

  • saying the App Store launch has already happened
  • implying GitHub and App Store shipped simultaneously when they did not

Product proof points

Use these when a GitHub release needs concrete detail:

  • evidence -> diagnosis -> prescription -> verification -> casebook
  • professional-role contract and role-drift labeling for named bots or seats
  • desktop hover companion that exposes supervision pressure before opening the full workspace
  • case-based remediation instead of generic monitoring-only alerts
  • template validation before broader promotion

Launch-safe wording

Preferred wording:

  • Apple-native correction workspace for supervising drifting bots
  • professional-role correction companion for OpenClaw
  • evidence-first intervention and casebook learning
  • correction-first supervision surface

Avoid:

  • claiming universal market exclusivity as a hard fact
  • implying that VeriClaw replaces all observability tooling
  • implying that the current launch ships watchOS as part of the release window
  • implying fully autonomous correction when user confirmation still matters

SEO and distribution hooks

Use these phrases naturally across the README, landing page, release body, and forwardable blurbs:

  • AI agent correction
  • LLM supervision
  • hallucination remediation
  • role-drift diagnosis
  • macOS-native OpenClaw companion
  • evidence -> diagnosis -> intervention -> verification

These phrases should stay attached to concrete product claims rather than becoming a loose keyword pile.

Download and share CTA lines

Use these when you want stronger click-through into the release or repository:

  • Open the GitHub-first launch pack for VeriClaw 爪印
  • Download the source package and forward the launch kit
  • Share the release page, screenshots, and source-availability guidance
  • Star or watch the repo to track the public launch surface

FAQ copy

What is VeriClaw?

  • VeriClaw 爪印 is the native correction companion for OpenClaw. It turns evidence of bot drift into diagnosis, recommended intervention, verification, and casebook learning.

Is VeriClaw separate from OpenClaw?

  • It is a companion, not a replacement. OpenClaw remains the runtime and gateway. VeriClaw adds the Apple-native supervision and correction layer.

Is this just another AI monitor?

  • No. Monitoring is part of the workflow, but the product is designed around correction. The core loop is evidence, diagnosis, prescription, verification, and casebook learning.

What makes it different from generic agent observability?

  • The product focuses on professional-role drift, case-based remediation, and native supervision surfaces rather than only dashboards, traces, or remote terminals.

Is watchOS part of this launch?

  • No. Watch is outside the current release gate.

Is the App Store live right now?

  • Not yet. The current public launch surface is GitHub first while the Apple companion continues through final hardening and submission steps.

Can people inspect the source freely?

  • They can inspect the public source for evaluation, citation, and internal review. Commercial use, public modified redistribution, and brand use require permission.

Screenshot caption direction

Suggested GitHub screenshot or media caption style:

  1. See drift pressure before it becomes invisible
  2. Open a case, not just a log
  3. Diagnose role drift and choose the next intervention
  4. Verify the fix before calling the loop closed