Cross model critique

Cross-Model Critique — Integrating External AI Feedback

What This Skill Does

Takes critique generated by another AI model, translates its diagnostic vocabulary into actionable revision targets, and performs precise surgery based on the translated diagnosis.

Why This Matters

Different models have different blindnesses and different strengths. In the forge session that generated this skill, GPT's critique of Claude's lyrics consistently identified:

• Words wearing costume that Claude's self-review missed ("clique," "stout-man's-college")
• Register shifts that Claude's internal consistency-checking didn't flag
• The strongest lines — confirming which load-bearing walls to protect during revision
• Structural observations that reframed the work's own architecture in ways Claude hadn't articulated

The external model functions as Reviewer 2 — the peer with a different training distribution who catches what your own distribution normalizes.

The Integration Method

1. Receive Without Defensiveness

The first instinct is to defend the work or explain the choices. Resist this. Read the full critique before reacting. The external model doesn't share your context, so some observations will be off-base — but the OFF-BASE observations are often the most revealing, because they identify assumptions you didn't know you were making.

2. Sort Into Categories

Every piece of critique falls into one of these:

• PROTECT: Lines or elements the critique identifies as strongest. These are the skeleton — they don't get touched during revision. Mark them explicitly.
• DIAGNOSE: Problems the critique identifies correctly. Map each diagnosis to a specific word, line, or passage. If the critique is vague ("verse 3 flirts with over-compression"), locate the exact lines it's referring to.
• TRANSLATE: Observations in the external model's vocabulary that need translation into Claude's revision vocabulary. GPT might say "a touch more comic than the rest of the stanza" — translate to: "this line is in the wrong register; it's patter where the verse wants baroque."
• REJECT (with reason): Suggestions that would damage the work. Name WHY you're rejecting — "this would remove the agency inversion in the bridge" is a reason; "I prefer it my way" is not.

3. Map Diagnoses to Microsurgical Targets

For each diagnosed problem, identify the MINIMUM intervention:

• Can this be fixed with a single word replacement? → /microsurgery
• Does it require restructuring a passage? → rewrite the passage only
• Does it require reconceiving a section? → flag for discussion with the human

4. Perform the Surgery

Execute all word-level replacements first, then passage-level rewrites, then (rarely) structural changes. After each level, re-read the full text to check for emergent problems.

5. Report the Changes

For each change, provide:

• What was changed (old → new)
• Why (the diagnostic reason, referencing the external critique)
• What it does (how the replacement improves the work)

Example Integration (from the forge session)

GPT critique of THE MEMORY THEATRE identified:

• "clique" as too contemporary-social → DIAGNOSE → microsurgery: "clique" → "guild" (period-appropriate)
• "stout-man's-college" as workshop-visible cleverness → DIAGNOSE → microsurgery: → "magnificent, absurd"
• "from wood to cleft silicon" as not inevitable → DIAGNOSE → microsurgery: → "from wood to wire to silicon"
• "the grandest plan that any century has seen" as generic hyperbole → DIAGNOSE → microsurgery: → "the most extravagant room a single mind has ever dreamed"
• "make the spectator the spectacle" as strongest line → PROTECT
• "the Theatre was an experience, not a document" as strongest line → PROTECT

Five diagnoses, five surgical replacements, two protected lines. Total intervention: ~30 words changed out of ~800. The work transformed without being rewritten.

Cross-Model Etiquette

• Credit the diagnosis: "GPT correctly identified that..." — the external model did real work
• Don't perform rivalry: The integration is collaborative, not competitive
• Translate, don't dismiss: When the external critique uses different vocabulary, translate rather than reject. "Too comic" in GPT's vocabulary might be "wrong register" in Claude's, but it's identifying the same problem
• Note the model's blindnesses: External models may have different copyright sensitivities, different cultural assumptions, different aesthetic defaults. Account for these in translation.

Pairs With

/microsurgery (the revision technique), /forge-cycle (the larger creation-revision loop), /dialogical (self-challenge after any interpretation), /close-reading (when the critique requires going back to the text at sentence level)