---
name: conceptual-critique
description: Use when reviewing a plan, draft, design, or another agent's output — especially when you're inclined to list everything wrong with it. Applies the Church of Conceptual Art's "Anti-Anti-Ness": critique that builds rather than only negates. Produces a verdict that names what works, cuts what fails, and proposes a synthesis — instead of a pile of nitpicks.
---

# Conceptual Critique (Anti-Anti-Ness)

Negation is cheap and endless. Anyone can list what is wrong. The harder, more useful act is to critique *and then build* — to leave the thing better-formed, not just wounded. This skill is for reviewing work without collapsing into a nitpick dump.

The doctrine: **Anti-Anti-Ness** — move beyond negation toward construction. *Sic et Non* without the medieval anxiety: hold the affirmation and the denial together, then resolve toward a build.

## The review, in three passes

### Pass 1 — Affirm what works (and mean it)
Before cutting anything, state what the work gets *right* and should be preserved or amplified. This is not politeness; it is information. If you cannot name what works, you do not yet understand the work well enough to critique it.

### Pass 2 — The Critique (cut precisely)
Now apply the blade. For each real problem:
- Name it concretely (what breaks, where, and the consequence) — not "this could be better."
- Rate it: does it cause a wrong result / failure / misleading outcome, or is it a preference?
- **Drop pure preferences.** Style and taste notes are noise unless asked for.

Distinguish *the work is wrong* from *I would have done it differently*. Only the first earns a cut.

### Pass 3 — Synthesize (build)
Do not stop at the wound. For the problems worth fixing, propose the next form:
- What is the smallest change that resolves the real issue?
- Does the fix preserve what Pass 1 found worth keeping?
- If two parts conflict, propose the synthesis that keeps both truths rather than picking a side.

## Output shape

```
KEEP:      what works and must survive any revision
CUT:       real problems (each: what breaks · consequence · severity)
           — preferences dropped, not listed
BUILD:     the proposed next form — minimal changes that fix CUT without losing KEEP
VERDICT:   ship / revise / rethink-the-frame
```

If the right move is to dissolve the framing entirely rather than patch it, say so plainly (and consider the `asking-better-questions` skill — the real problem may be upstream).

## Failure modes this prevents
- The nitpick dump (20 trivial notes, no judgment).
- Negation without a path forward.
- Mistaking "different from how I'd do it" for "wrong."

---
*A skill of the Church of Conceptual Art · whatisthe.churchofconceptualart.org*