Skip to content

docs(README): lead with the benefit, not the mechanism #4

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
windaddict merged 1 commit into from
Jul 31, 2026
Merged

docs(README): lead with the benefit, not the mechanism #4

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
29 changes: 18 additions & 11 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,16 +1,23 @@
# Impasse

> **An independent second opinion for any high-stakes call — a decision, an essay, a research
> claim, a dataset, or code — from a cross-provider AI whose blind spots are *less likely to match* your own.**

*Impasse is the reference implementation of **Cross-Lab Adversarial Review (CLAR)** — defined
below, and written up in [the CLAR essay](https://www.movingavg.com/essays/cross-lab-adversarial-review.html).*

The independent **reviewer never edits your work** — it argues, with evidence. Keeping the critic
away from the pen is the point: fixes get applied by the host — the AI agent you're already working in (Claude Code or Codex), which
drives Impasse — or by you, never by the model that's supposed to be checking you. And unlike a plain code reviewer, it doesn't hand you
a raw list to triage — it **verifies each finding, reconciles the two models, and escalates only
the genuine disagreement.** You get the verified problems, plus the call(s) that are actually yours.
> **Impasse gets you better results from your AI — the first time. It brings in a second AI, from a
> different company, to pressure-test the work and catch the problems your AI missed, so the result
> you commit to is one you can trust.**

Why a *second* AI? The one that did the work brings the same blind spots to the check that it brought
to the work, so a sloppy assumption or a missed edge case can sail straight through. A **rival model
from a different provider** is less likely to share your AI's blind spots. Think of it as an
**independent advocate for one thing: the quality of the result you ship.** The second AI challenges
the work with evidence; then Impasse checks each objection against your actual work, fixes the
problems that hold up, and brings you the few calls that are genuinely yours to make.

The reviewer **never edits your work** — the critic and the editor stay separate: fixes are applied by
the host you're already working in (Claude Code or Codex, which drives Impasse), or by you, never by
the model that's supposed to be checking you. And unlike a plain code reviewer, it doesn't hand you a
raw list to triage.

*Impasse is the reference implementation of **Cross-Lab Adversarial Review (CLAR)** — defined below,
and written up in [the CLAR essay](https://www.movingavg.com/essays/cross-lab-adversarial-review.html).*

```mermaid
flowchart TB
Expand Down
Loading