review(pr-1255-postmerge): row-7 prose-form fix + drift catalogue extended to 15 instances

[AceHack]

Summary

P2 finding on PR #1255 (already merged): my "fix" of row 7 in the skill-design memo was a second semantic-equivalence drift on the same substitution attempt.

Two recursive failures on the same substitution:

1. First attempt: ls docs/DECISIONS/ | grep -iE → find docs/DECISIONS/ -iname "*<pattern>*" (caught: find -iname only does shell glob, not regex alternation)
2. Second attempt: → grep -ilrE "<pattern>" docs/DECISIONS/ (caught: grep -r searches file CONTENTS, not filenames)

Both attempts failed because I claimed semantic-equivalence without empirical verification. Exactly the failure mode the verify-then-claim discipline names; happening on the very memo defining the discipline.

Fix

• Row 7 of the skill-design memo's procedure-table now describes the search in prose with a NOTE warning against the wrong substitution
• Canonical command pulled OUT of the table cell into a fenced code block below the table — preserves working command without pipe-in-cell escapes

Drift catalogue extended

The verify-then-claim memo's drift catalogue grew from 9 → 15+ instances. New entries (post-naming):

• Round 33 followup — bash is Unix-only; cross-platform automation = TypeScript/Bun #10-Round 33 — check in .vscode/, reframe post-install automation as research #11: self-recursive drift (memo about drift contains its own drift)
• Round 33 — backlog scope audit + GOVERNANCE §29 #12-Round 33 — docs/VISION.md first draft + product-visionary role proposal #13: same substitution attempt failed twice (find-iname + grep-ilrE)
• Round 33 — VISION.md v2: Zeta is a full database, factory is 1st-class #14: convention drift (ADR supersession marker)
• Round 33 — VISION.md v3: SQL frontend in v1, Reaqtor niche identified, 5 answers locked #15: path-form drift (fully-qualified vs bare paths inconsistent)

Added new section identifying recurring sub-classes:

• existence-drift / count-drift / semantic-equivalence-drift / empirical-output-drift / convention-drift / path-form-drift / self-recursive-drift

Headline update: 6 of 15 instances landed AFTER the discipline was named. Manual discipline insufficient. tools/substrate-claim-checker/ TS tool is the only path forward.

Test plan

• Row 7 description prose-form, no pipe-in-cell
• Canonical ls | grep -iE command in fenced code block below the table
• Verify-then-claim catalogue extended with instances Round 33 followup — bash is Unix-only; cross-platform automation = TypeScript/Bun #10-Round 33 — VISION.md v3: SQL frontend in v1, Reaqtor niche identified, 5 answers locked #15
• Sub-classes section added
• MEMORY.md not modified (no new memo file added; only existing memo updated)
• CI green

🤖 Generated with Claude Code

[Copilot]

Pull request overview

Updates two existing memory/ feedback memos to correct a previously-merged semantic-equivalence drift in an ADR-filename search procedure and to expand the “verify-then-claim” drift catalogue/taxonomy with additional instances.

Changes:

• Rewrites the skill-design memo’s row-7 ADR search guidance into prose and moves the canonical ls | grep -iE example into a fenced code block (avoids markdown table pipe escaping and warns against the incorrect grep -r substitution).
• Extends the verify-then-claim memo’s empirical drift catalogue from 9→15+ instances and adds a recurring sub-classes breakdown.
• Updates surrounding narrative to reflect the new drift instances and the recursive failure mode.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 3 comments.

File	Description
memory/feedback_verify_then_claim_discipline_dominant_failure_mode_substrate_authoring_otto_2026_05_03.md	Expands drift-instance catalogue and adds drift sub-class taxonomy.
memory/feedback_skills_as_carved_sentences_knowledge_in_docs_datavault_2_0_pattern_aaron_2026_05_03.md	Fixes row-7 ADR-filename search guidance and adds canonical command as a fenced block.

[AceHack]

All 3 post-merge findings addressed in follow-up PR #1259:

• Frontmatter description count drift (9→18+) → fixed
• Body line 91 count drift → fixed
• MEMORY.md index entry count drift + manual-insufficient framing → updated

This is itself drift instance #19 (frontmatter ↔ body sub-pattern within count drift). Both substrate surfaces should match body content; sub-class added to substrate-claim-checker spec.