Skip to content

Latest commit

 

History

History
24 lines (21 loc) · 2.16 KB

ALIGNMENT.md

File metadata and controls

24 lines (21 loc) · 2.16 KB

ALIGNMENT — R13 compliance ledger (dead-code-quantifier)

Date: 2026-10-11 · Round: R13 · Worker: W6 · Status: aligned

# Requirement Status Evidence
1 Pure Python standard library, zero third-party deps PASS imports: argparse/ast/json/os/sys only
2 Runnable CLI with argparse entry PASS python dead_code_quantifier.py scan <targets> with --mode/--format/--gate
3 self_test.py ≥40 cases PASS round 1 = 40 checks, round 2 +10 → 50 checks
4 Output ends with SUMMARY: PASS n/n PASS final print in self_test.py
5 Three test categories PASS blocks A (1–20), E (21–30), X (31–40), plus round-2 R block
6 Edge classes: empty / very long / corrupted / Unicode / deep nesting PASS E02 empty, X01 502-loc def, E01/E08 corrupted, X02 unicode names, X08 300-level nesting
7 README first paragraph ≤60 words PASS 40 words (self-counted, verified via sed -n '3p' README.md | wc -w)
8 why-not-X uses real anchors; sim: none → exact sentence PASS README states "No direct equivalent found as of 2026-10-11" + 3 adjacent tools with layer differences (design JSON anchors: [])
9 Minimal Fallback section PASS README §Minimal Fallback (grep reference count)
10 Design Trade-offs section PASS README §Design Trade-offs (4 items)
11 Limits & Second-order Effects PASS README §Limits (dynamic dispatch blindness, tool-disabling second-order effect)
12 FAQ incl. When NOT to use PASS README §FAQ final item
13 [Core Assumptions] ≥3 PASS README §Core Assumptions (4 items)
14 MIT LICENSE "zhengming contributors", no real names/keys PASS LICENSE line 3; no secrets/real names anywhere
15 CHANGELOG v0.1.0 → v0.1.1 revision trail PASS CHANGELOG.md, both entries dated 2026-10-11
16 Number discipline: measured with repro or [estimated] PASS all timings tagged [estimated] with repro commands

Two-iteration record: draft → self-review found aggressive mode untested for lambda/walrus sources and no gate boundary tests → round 2 added 10 checks (+25% tests) covering those. Tests re-run green after round 2.