REMORSE builds agent products on Claude. Remorsearch lets Claude research, test and fix the UX of real products, with evidence for every claim. Maccept brings your own Claude Code to Siri on your Mac.
> /remorsearch-ux-qa Improve the checkout UX URL: https://shop.example.test ● Examine product, docs, recorded decisions ● Test persona: first-time buyer (synthetic) [observed] Card form resets after a failed payment [inferred] Buyers may drop off at re-entry [unknown] Share of mobile buyers (no data) ● Research 4 public sources · 2 coverage gaps ● Fix keep valid fields (reversible) ● Verify replay passed · no regression >
Every result says what was observed, what was inferred and what is still unknown.
An Agent Skill for Claude Code. Give Claude a product URL or a repository. It examines the product, runs persona QA, researches real user experiences, applies reversible fixes and replays the tasks to verify them.
Explore Remorsearch →Siri for your own Claude Code. Start, continue, correct or stop work by talking to Siri on your Mac, and get the result inside Siri. The Claude Code connection is implemented and in live verification.
See Maccept →Claude reads the product and the recorded decisions of its developers, so intentional choices stay intact.
Baseline and persona QA: task completion, input preservation, error recovery, keyboard use, KWCAG and WCAG 2.2.
Real user experiences from permitted public sources, with provenance and personal data removed.
Scoped, reversible fixes. Screen changes go to Figma first, with the file, node and version on record.
A replay on the current build, an adjacent regression check and an intent check. Only then is it done.
Our products give Claude a workflow, guardrails and verification. Claude reads the product, plans, uses tools and writes the fixes.
Every statement is labeled observed, inferred or unknown. Synthetic personas create test conditions. They never stand in for real users.
Changes stay in scope and can be reverted. An improvement counts only after a replay on the current build.
Remorsearch ships as an Agent Skill with a Claude Code plugin manifest. A team installs it by copying one folder.
Skills also run in headless Claude Code (claude -p). Maccept connects Siri to your own Claude Code in headless mode.
Our tools run on your machine. We do not operate servers that receive your product data. Research removes personal data before Claude reads it.
Self-improvement of an autonomous optimization research loop, with no human in the loop.
Regression leaks across more than 250 recorded runs of that loop.
Merged pull requests from pre-registered, blind-evaluated agent-delegation experiments across Claude models.
Entrants at AX Talent War 2026, where the founder was named a top AI talent and reserve finalist.
October 2026. MIT license. Install it from GitHub into Claude Code or Codex.
The Claude Code connection is implemented and covered by unit tests. Live verification on real Siri is in progress.
We will share release details on the Maccept page.
We are looking for design partners for Remorsearch and early users for Maccept.
REMORSE