Skip to content

review-ledger: enforce the entry template with an executable lint (Ideal prompt diff / What was fucked up) #640

Description

@4444J99

Fable adjudication of Codex session 019f2a04 (full-stack review founding run) found template drift in docs/agent-code-diff-review.md: of 182 entries in the ledger at the session's last commit (f404f85), only 124 carry the literal Ideal prompt diff: section and 123 carry What was fucked up: (182/182 have Severity+Evidence; 179 have Verification+Result). Those two lenses are the mandate's signature deliverable ("the ideal forms diff, not just the diff of code") and must be structurally enforced, not habitual.

Done predicate: a lint script (e.g. scripts/check-review-ledger.py, wired into scripts/verify-whole.sh or the doc-writing loop) exits 0 ⟺ every ### entry in docs/agent-code-diff-review.md contains the required sections: Severity, Evidence, Ideal prompt diff, Outcome, What was fucked up, Verification, Result. Existing non-conforming entries are either backfilled or explicitly grandfathered by a dated allowlist inside the script.

Counts verified via: git show f404f85:docs/agent-code-diff-review.md | grep -oE '^(Severity|Evidence|Ideal prompt diff|Outcome|What was fucked up|Verification|Result):' | sort | uniq -c

🤖 Generated with Claude Code

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Fields

    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions