1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566 | #!/usr/bin/env python3
"""
verify_copywriting.py — the anti-theater gate for the master-copywriter harness.
A REAL linter (python3 stdlib only) that scores ONE deliverable file against
the Conversion Copywriting rubric in MANUAL.md §8. Exits 0 if the deliverable
meets the standard, 1 if it fails — always printing which named checks failed.
Usage:
python3 verify_copywriting.py <type> <file>
<type> is one of: sales-page | vsl-script | ad-set | email-subject-bank | landing-hero-variants
This is deliberately a linter, not a vibe check: every rule below is
mechanically checkable and traces to a named §8 rubric line in MANUAL.md.
"""
from __future__ import annotations
import re
import sys
# ---------------------------------------------------------------------------
# Shared vocabulary gates (stand-in for the `brand-bible` skill's live list —
# sync these against the skill's forbidden-vocabulary list when it is
# available in the running environment; kept local here so this linter has
# zero runtime dependencies).
# ---------------------------------------------------------------------------
HYPE_WORDS = [
"guaranteed riches", "effortless", "overnight", "get rich quick",
"life-changing money", "instant wealth", "no work required",
"risk-free money", "secret loophole", "magic formula",
"revolutionary", "insane results", "too good to be true",
]
FORBIDDEN_VOCAB = [
"delve into", "in today's fast-paced world", "unlock the power",
"game-changer", "game changing", "supercharge", "unleash your potential",
"take it to the next level", "in this day and age", "at the end of the day",
"synergy", "leverage synergies", "circle back",
]
PROOF_TAGS = ("testimonial", "data", "demo", "authority", "story")
CURIOSITY_MECHANISMS = ("open loop", "specificity", "contrast", "self-interest", "story-cue")
LEAD_TYPES = ("offer", "promise", "problem-solution", "big-secret", "proclamation", "story")
AWARENESS_STAGES = ("unaware", "problem-aware", "solution-aware", "product-aware", "most-aware")
SPAM_TRIGGER_PATTERNS = [
re.compile(r"free!!!+", re.I),
re.compile(r"\bact now\b", re.I),
re.compile(r"\${2,}", re.I),
]
EMOJI_PATTERN = re.compile(
"["
"\U0001F300-\U0001FAFF"
"\U00002600-\U000027BF"
"\U0001F1E6-\U0001F1FF"
"]"
)
class Report:
"""Collects named check results; exit code reflects whether any failed."""
def __init__(self, label: str):
self.label = label
self.failures: list[str] = []
self.passes: list[str] = []
def check(self, name: str, ok: bool, detail: str = ""):
line = f"{name}" + (f" — {detail}" if detail else "")
if ok:
self.passes.append(line)
else:
self.failures.append(line)
def render_and_exit(self) -> int:
print(f"== verify_copywriting :: {self.label} ==")
for p in self.passes:
print(f" PASS: {p}")
for f in self.failures:
print(f" FAIL: {f}")
if self.failures:
print(f"RESULT: FAIL ({len(self.failures)} check(s) failed, {len(self.passes)} passed)")
return 1
print(f"RESULT: PASS ({len(self.passes)} check(s) passed)")
return 0
# ---------------------------------------------------------------------------
# Text-level primitives shared across deliverable types
# ---------------------------------------------------------------------------
def word_count(text: str) -> int:
return len(re.findall(r"[A-Za-z0-9'’]+", text))
def count_syllables(word: str) -> int:
word = word.lower().strip(".,!?;:\"'()[]")
if not word:
return 0
word = re.sub(r"[^a-z]", "", word)
if not word:
return 1
groups = re.findall(r"[aeiouy]+", word)
count = len(groups)
if word.endswith("e") and not word.endswith("le") and count > 1:
count -= 1
return max(count, 1)
def flesch_reading_ease(text: str) -> float:
"""Rudolf Flesch's Reading Ease formula, computed with stdlib only."""
sentences = [s for s in re.split(r"[.!?]+", text) if s.strip()]
words = re.findall(r"[A-Za-z']+", text)
if not sentences or not words:
return 0.0
syllables = sum(count_syllables(w) for w in words)
n_words = len(words)
n_sentences = len(sentences)
return 206.835 - 1.015 * (n_words / n_sentences) - 84.6 * (syllables / n_words)
def paragraphs(text: str) -> list[str]:
"""Split into prose 'paragraphs' for the 90-word gate.
Blank-line-separated blocks are the unit, EXCEPT a block that is made up
entirely of list-marker lines (`- ...` / `1. ...`) is a list, not a
paragraph — each list item is checked on its own so a 4-item bullet list
doesn't get merged into one giant "paragraph" and unfairly fail the gate.
"""
out: list[str] = []
for block in re.split(r"\n\s*\n", text):
block = block.strip()
if not block:
continue
lines = [l for l in block.splitlines() if l.strip()]
is_list_block = bool(lines) and all(re.match(r"^\s*(?:[-*]|\d+[.)])\s+", l) for l in lines)
if is_list_block:
out.extend(lines)
else:
out.append(block)
return out
def em_dash_density_per_200(text: str) -> float:
dashes = text.count("—") + text.count("--")
words = word_count(text)
if words == 0:
return 0.0
return dashes / (words / 200.0)
def scan_banned(text: str, wordlist: list[str]) -> list[str]:
hits = []
low = text.lower()
for w in wordlist:
if w.lower() in low:
hits.append(w)
return hits
def find_section(text: str, name: str):
"""Return (start_index, header_match_end) for a `## Name` header, or None."""
m = re.search(rf"^##\s+{re.escape(name)}\b.*$", text, re.I | re.M)
return m
def section_body(text: str, headers_in_order: list[str], name: str) -> str:
"""Slice the body of section `name` up to the next known header (or EOF)."""
idx = headers_in_order.index(name)
start_m = find_section(text, name)
if not start_m:
return ""
start = start_m.end()
end = len(text)
for later in headers_in_order[idx + 1:]:
m = find_section(text, later)
if m:
end = min(end, m.start())
return text[start:end]
# ---------------------------------------------------------------------------
# sales-page (MANUAL.md §8.1)
# ---------------------------------------------------------------------------
def verify_sales_page(text: str, r: Report):
order = ["Headline", "Lede", "Mechanism", "Proof", "Offer", "Objections", "CTA", "PS"]
positions = {}
for name in order:
m = find_section(text, name)
positions[name] = m.start() if m else None
missing = [n for n in order if positions[n] is None]
r.check("all 8 required sections present (Headline/Lede/Mechanism/Proof/Offer/Objections/CTA/PS)",
not missing, f"missing: {', '.join(missing)}" if missing else "")
if not missing:
seq = [positions[n] for n in order]
in_order = all(seq[i] < seq[i + 1] for i in range(len(seq) - 1))
r.check("sections appear in required order", in_order,
"" if in_order else "headers are present but out of order")
else:
in_order = False
# Headline block: lead-type tag + awareness tag
headline_body = section_body(text, order, "Headline") if positions["Headline"] is not None else ""
lead_m = re.search(r"\[LEAD:\s*([^\]]+)\]", headline_body, re.I)
aware_m = re.search(r"\[AWARENESS:\s*([^\]]+)\]", headline_body, re.I)
lead_ok = bool(lead_m) and lead_m.group(1).strip().lower() in LEAD_TYPES
aware_ok = bool(aware_m) and aware_m.group(1).strip().lower() in AWARENESS_STAGES
r.check("headline block declares a valid lead-type tag (one of the six Great Leads)",
lead_ok, f"found: {lead_m.group(1) if lead_m else 'none'}")
r.check("headline block declares a valid awareness-stage tag",
aware_ok, f"found: {aware_m.group(1) if aware_m else 'none'}")
# Mechanism: a named mechanism string
mech_body = section_body(text, order, "Mechanism") if positions["Mechanism"] is not None else ""
mech_m = re.search(r"Mechanism:\s*(.+)", mech_body, re.I)
mech_ok = bool(mech_m) and len(mech_m.group(1).strip()) >= 3
r.check("mechanism section names a mechanism string (`Mechanism: <name>`)",
mech_ok, "" if mech_ok else "no `Mechanism:` line with a real name found")
# Proof block: >=3 tagged elements
proof_body = section_body(text, order, "Proof") if positions["Proof"] is not None else ""
proof_lines = [l for l in proof_body.splitlines() if l.strip().startswith("-")]
tag_re = re.compile(r"\[(" + "|".join(PROOF_TAGS) + r")\]", re.I)
tagged = [l for l in proof_lines if tag_re.search(l)]
r.check("proof block has >=3 tagged proof elements", len(tagged) >= 3,
f"found {len(tagged)} tagged proof line(s)")
untagged_with_numbers = [l for l in proof_lines if not tag_re.search(l) and re.search(r"\d", l)]
r.check("no untagged proof-block claims carrying numbers", not untagged_with_numbers,
f"{len(untagged_with_numbers)} untagged numeric claim(s) in proof block")
# Untagged numeric claims elsewhere (lede / mechanism / objections) must not
# smuggle in bare stats without a proof tag on the same line.
other_sections_text = "\n".join(
section_body(text, order, n) for n in ("Lede", "Mechanism", "Objections") if positions.get(n) is not None
)
stat_line_re = re.compile(r"^[^\n]*\d[^\n]*%[^\n]*$|^[^\n]*\$\d[^\n]*$", re.M)
smuggled = [l for l in stat_line_re.findall(other_sections_text) if not tag_re.search(l)]
r.check("no untagged numeric/statistical claims smuggled outside the proof block",
not smuggled, f"{len(smuggled)} untagged stat line(s) found outside Proof")
# Offer: price, stack, guarantee
offer_body = section_body(text, order, "Offer") if positions["Offer"] is not None else ""
has_price = bool(re.search(r"Price:\s*\$?\d", offer_body, re.I))
has_stack = bool(re.search(r"Stack:", offer_body, re.I)) and len(
[l for l in offer_body.splitlines() if l.strip().startswith("-")]) >= 1
has_guarantee = bool(re.search(r"Guarantee:\s*\S+", offer_body, re.I))
r.check("offer section has a Price:", has_price)
r.check("offer section has a Stack: with items", has_stack)
r.check("offer section has Guarantee: text", has_guarantee)
# Objections: >=3
obj_body = section_body(text, order, "Objections") if positions["Objections"] is not None else ""
obj_items = re.findall(r"^\s*(?:\d+[.)]|-)\s+\S", obj_body, re.M)
r.check("objection handling has >=3 objections", len(obj_items) >= 3,
f"found {len(obj_items)}")
# CTA: exactly one distinct target
cta_body = section_body(text, order, "CTA") if positions["CTA"] is not None else ""
cta_matches = re.findall(r"CTA:\s*\[[^\]]*->\s*([^\]]+)\]", cta_body, re.I)
distinct_targets = {t.strip() for t in cta_matches}
r.check("exactly one distinct CTA target declared", len(distinct_targets) == 1,
f"found {len(distinct_targets)} distinct target(s): {sorted(distinct_targets)}")
# PS block
ps_body = section_body(text, order, "PS") if positions["PS"] is not None else ""
r.check("PS block has content", bool(ps_body.strip()))
# Banned patterns (hard fails, whole document)
hype_hits = scan_banned(text, HYPE_WORDS)
r.check("no hype-list words (guaranteed riches / effortless / overnight / ...)",
not hype_hits, f"hit(s): {hype_hits}" if hype_hits else "")
vocab_hits = scan_banned(text, FORBIDDEN_VOCAB)
r.check("no brand-bible forbidden vocabulary", not vocab_hits,
f"hit(s): {vocab_hits}" if vocab_hits else "")
long_paragraphs = [p for p in paragraphs(text) if word_count(p) > 90]
r.check("no paragraph exceeds 90 words", not long_paragraphs,
f"{len(long_paragraphs)} paragraph(s) over 90 words")
dash_density = em_dash_density_per_200(text)
r.check("em-dash density <=1 per 200 words", dash_density <= 1.0,
f"{dash_density:.2f} per 200 words")
fre = flesch_reading_ease(text)
r.check("Flesch Reading Ease >= 50", fre >= 50, f"scored {fre:.1f}")
# ---------------------------------------------------------------------------
# vsl-script (MANUAL.md §8.2)
# ---------------------------------------------------------------------------
def verify_vsl_script(text: str, r: Report):
order = ["Hook", "Lead", "BigIdea", "Proof", "Transition", "Stack", "PriceReveal", "Guarantee", "CTA"]
aliases = {
"Hook": "Hook", "Lead": "Lead", "BigIdea": "Big Idea",
"Proof": "Proof", "Transition": "Transition", "Stack": "Stack",
"PriceReveal": "Price Reveal", "Guarantee": "Guarantee", "CTA": "CTA",
}
positions = {}
for key, display in aliases.items():
m = find_section(text, display)
positions[key] = m.start() if m else None
missing = [aliases[k] for k in order if positions[k] is None]
r.check("all 9 required beats present (Hook/Lead/Big Idea/Proof/Transition/Stack/Price Reveal/Guarantee/CTA)",
not missing, f"missing: {', '.join(missing)}" if missing else "")
display_order = [aliases[k] for k in order]
def body_of(key):
return section_body(text, display_order, aliases[key]) if positions[key] is not None else ""
hook_body = body_of("Hook")
ts_ok = bool(re.search(r"\[00:0[0-9]\s*-\s*00:1[0-5]\]|\[0:00\s*-\s*0:15\]|\[first 15s\]", hook_body, re.I))
r.check("hook has a first-15s timestamped block", ts_ok,
"" if ts_ok else "no [0:00-0:15]-style timestamp found")
hook_text_m = re.search(r"Hook:\s*\"([^\"]+)\"", hook_body, re.I)
hook_words = word_count(hook_text_m.group(1)) if hook_text_m else 999
r.check("hook line is <=40 words", bool(hook_text_m) and hook_words <= 40,
f"{hook_words} words" if hook_text_m else "no `Hook: \"...\"` line found")
lead_body = body_of("Lead")
lead_m = re.search(r"\[LEAD:\s*([^\]]+)\]", lead_body, re.I)
r.check("lead beat declares a valid lead-type tag",
bool(lead_m) and lead_m.group(1).strip().lower() in LEAD_TYPES,
f"found: {lead_m.group(1) if lead_m else 'none'}")
bigidea_body = body_of("BigIdea")
r.check("big-idea/mechanism beat names a mechanism",
bool(re.search(r"Mechanism:\s*\S+", bigidea_body, re.I)))
proof_body = body_of("Proof")
proof_lines = [l for l in proof_body.splitlines() if l.strip().startswith("-")]
tag_re = re.compile(r"\[(" + "|".join(PROOF_TAGS) + r")\]", re.I)
tagged = [l for l in proof_lines if tag_re.search(l)]
r.check("proof segments >=3, all tagged", len(tagged) >= 3 and len(tagged) == len(proof_lines),
f"{len(tagged)}/{len(proof_lines)} tagged")
r.check("transition-to-pitch marker present", positions["Transition"] is not None)
r.check("stack beat present", positions["Stack"] is not None and bool(body_of("Stack").strip()))
price_body = body_of("PriceReveal")
anchor_ok = bool(re.search(r"(anchor|value)\D*\$\d", price_body, re.I)) and bool(re.search(r"Price:\s*\$?\d", price_body, re.I))
r.check("price reveal includes an anchor value and the price", anchor_ok)
r.check("guarantee beat present with text", bool(body_of("Guarantee").strip()))
cta_body = body_of("CTA")
cta_matches = re.findall(r"CTA:\s*\[[^\]]*->\s*([^\]]+)\]", cta_body, re.I)
distinct_targets = {t.strip() for t in cta_matches}
r.check("single CTA action (repetition of same target allowed)", len(distinct_targets) == 1,
f"found {len(distinct_targets)} distinct target(s)")
loops = re.findall(r"\[LOOP-OPEN:[^\]]+\]", text, re.I)
closes = re.findall(r"\[LOOP-CLOSE:[^\]]+\]", text, re.I)
def loop_id(tag):
m = re.search(r"\[LOOP-OPEN:\s*([^\]]+)\]|\[LOOP-CLOSE:\s*([^\]]+)\]", tag, re.I)
return (m.group(1) or m.group(2)).strip() if m else None
open_ids = {loop_id(t) for t in loops}
close_ids = {loop_id(t) for t in closes}
unclosed = open_ids - close_ids
r.check("open-loop inventory has >=2 loops, each with a declared close point",
len(open_ids) >= 2 and not unclosed,
f"{len(open_ids)} opened, unclosed: {sorted(unclosed)}" if (len(open_ids) < 2 or unclosed) else "")
# ---------------------------------------------------------------------------
# ad-set (MANUAL.md §8.3)
# ---------------------------------------------------------------------------
def verify_ad_set(text: str, r: Report):
variants = re.split(r"^##\s+Variant\s+\d+.*$", text, flags=re.M | re.I)[1:]
r.check(">=5 ad variants present", len(variants) >= 5, f"found {len(variants)}")
hooks = []
angle_counts: dict[str, int] = {}
problems = []
for i, v in enumerate(variants, start=1):
aware_m = re.search(r"\[AWARENESS:\s*([^\]]+)\]", v, re.I)
hook_m = re.search(r"Hook:\s*\"([^\"]+)\"", v, re.I)
cta_m = re.findall(r"CTA:\s*\[[^\]]*->\s*([^\]]+)\]", v, re.I)
angle_m = re.search(r"\[ANGLE:\s*([^\]]+)\]", v, re.I)
if not aware_m or aware_m.group(1).strip().lower() not in AWARENESS_STAGES:
problems.append(f"variant {i}: missing/invalid awareness tag")
if not hook_m:
problems.append(f"variant {i}: missing Hook: line")
else:
hooks.append(hook_m.group(1).strip().lower())
if word_count(hook_m.group(1)) > 12:
problems.append(f"variant {i}: hook line over 12 words")
if len({t.strip() for t in cta_m}) != 1:
problems.append(f"variant {i}: does not declare exactly one CTA target")
if not angle_m:
problems.append(f"variant {i}: missing [ANGLE: ...] tag")
else:
angle = angle_m.group(1).strip().lower()
angle_counts[angle] = angle_counts.get(angle, 0) + 1
income_claim = re.search(r"\b(before|after)\b.*\$", v, re.I)
disclaimer = re.search(r"disclaimer", v, re.I)
if income_claim and not disclaimer:
problems.append(f"variant {i}: before/after income claim without a disclaimer line")
dup_hooks = {h for h in hooks if hooks.count(h) > 1}
r.check("no duplicate hook lines across variants", not dup_hooks,
f"duplicated: {sorted(dup_hooks)}" if dup_hooks else "")
overused_angles = {a: c for a, c in angle_counts.items() if c > 2}
r.check("each angle tag used at most twice across the set", not overused_angles,
f"overused: {overused_angles}" if overused_angles else "")
r.check("every variant declares awareness tag, hook (<=12 words), single CTA, and angle tag",
not problems, "; ".join(problems) if problems else "")
# ---------------------------------------------------------------------------
# email-subject-bank (MANUAL.md §8.4)
# ---------------------------------------------------------------------------
def verify_email_subject_bank(text: str, r: Report):
groups = re.split(r"^##\s+(.+)$", text, flags=re.M)
# re.split with a capturing group interleaves headers and bodies
parsed = []
for i in range(1, len(groups), 2):
header, body = groups[i], groups[i + 1] if i + 1 < len(groups) else ""
parsed.append((header.strip(), body))
mechanism_hits = set()
all_subjects = []
problems = []
for header, body in parsed:
header_low = header.lower()
matched_mechanism = next((m for m in CURIOSITY_MECHANISMS if m in header_low), None)
cite_m = re.search(r"Mechanism:\s*(.+)", body, re.I)
if matched_mechanism:
mechanism_hits.add(matched_mechanism)
if not cite_m:
problems.append(f"group '{header}': missing `Mechanism:` citation line")
subj_lines = [l.strip()[2:].strip() for l in body.splitlines() if l.strip().startswith("-")]
for s in subj_lines:
all_subjects.append(s)
if word_count(s) > 9:
problems.append(f"subject over 9 words: '{s}'")
caps_words = [w for w in re.findall(r"[A-Za-z]+", s) if len(w) > 1 and w.isupper()]
if caps_words:
problems.append(f"ALL-CAPS word in subject: '{s}'")
emoji_count = len(EMOJI_PATTERN.findall(s))
if emoji_count > 1:
problems.append(f"more than 1 emoji in subject: '{s}'")
for pat in SPAM_TRIGGER_PATTERNS:
if pat.search(s):
problems.append(f"spam-trigger hit in subject: '{s}'")
r.check(">=20 subjects total", len(all_subjects) >= 20, f"found {len(all_subjects)}")
r.check(">=4 named curiosity mechanisms represented "
"(open loop/specificity/contrast/self-interest/story-cue)",
len(mechanism_hits) >= 4, f"found: {sorted(mechanism_hits)}")
r.check("every group cites its mechanism, and no subject fails length/caps/emoji/spam checks",
not problems, "; ".join(problems[:8]) + (" ..." if len(problems) > 8 else "") if problems else "")
# ---------------------------------------------------------------------------
# landing-hero-variants (MANUAL.md §8.5)
# ---------------------------------------------------------------------------
def verify_landing_hero_variants(text: str, r: Report):
variants = re.split(r"^##\s+Variant\s+\d+.*$", text, flags=re.M | re.I)[1:]
r.check(">=3 hero variants present", len(variants) >= 3, f"found {len(variants)}")
lead_types_seen = []
problems = []
for i, v in enumerate(variants, start=1):
headline_m = re.search(r"Headline:\s*(.+)", v, re.I)
subhead_m = re.search(r"Subhead:\s*(.+)", v, re.I)
cta_m = re.search(r"CTA:\s*(.+)", v, re.I)
lead_m = re.search(r"\[LEAD:\s*([^\]]+)\]", v, re.I)
aware_m = re.search(r"\[AWARENESS:\s*([^\]]+)\]", v, re.I)
because_m = re.search(r"Because:\s*(.+)", v, re.I)
if not headline_m:
problems.append(f"variant {i}: missing Headline:")
elif word_count(headline_m.group(1)) > 14:
problems.append(f"variant {i}: headline over 14 words")
if not subhead_m:
problems.append(f"variant {i}: missing Subhead:")
elif word_count(subhead_m.group(1)) > 25:
problems.append(f"variant {i}: subhead over 25 words")
if not cta_m:
problems.append(f"variant {i}: missing CTA:")
else:
cta_text = cta_m.group(1).strip()
cta_words = cta_text.split()
first_word = re.sub(r"[^A-Za-z]", "", cta_words[0]) if cta_words else ""
verb_first = first_word.lower().endswith(("t", "n", "m", "k", "d", "w", "y", "e", "p", "g", "r", "s", "l", "h", "b", "c", "f", "v", "x", "z", "a", "i", "u", "o", "q", "j"))
if len(cta_words) > 4:
problems.append(f"variant {i}: CTA label over 4 words")
if not cta_words or not first_word:
problems.append(f"variant {i}: CTA label unparsable")
if not lead_m or lead_m.group(1).strip().lower() not in LEAD_TYPES:
problems.append(f"variant {i}: missing/invalid lead-type tag")
else:
lead_types_seen.append(lead_m.group(1).strip().lower())
if not aware_m or aware_m.group(1).strip().lower() not in AWARENESS_STAGES:
problems.append(f"variant {i}: missing/invalid awareness tag")
if not because_m or not because_m.group(1).strip():
problems.append(f"variant {i}: missing 'Because:' rationale")
r.check("every variant has headline<=14w, subhead<=25w, verb-first CTA<=4w, "
"lead+awareness tags, and a Because: rationale", not problems,
"; ".join(problems[:8]) + (" ..." if len(problems) > 8 else "") if problems else "")
diversity_ok = len(set(lead_types_seen)) >= 2 if lead_types_seen else False
r.check("not all variants share the same lead type (diversity requirement)",
diversity_ok, f"lead types used: {lead_types_seen}")
# ---------------------------------------------------------------------------
# entry point
# ---------------------------------------------------------------------------
DISPATCH = {
"sales-page": verify_sales_page,
"vsl-script": verify_vsl_script,
"ad-set": verify_ad_set,
"email-subject-bank": verify_email_subject_bank,
"landing-hero-variants": verify_landing_hero_variants,
}
def main(argv: list[str]) -> int:
if len(argv) != 3:
print(f"usage: {argv[0]} <{'|'.join(DISPATCH)}> <file>", file=sys.stderr)
return 2
dtype, path = argv[1], argv[2]
if dtype not in DISPATCH:
print(f"FAIL: unknown deliverable type '{dtype}'. Valid: {', '.join(DISPATCH)}", file=sys.stderr)
return 2
try:
with open(path, "r", encoding="utf-8") as f:
text = f.read()
except OSError as e:
print(f"FAIL: cannot read {path}: {e}", file=sys.stderr)
return 1
r = Report(f"{dtype} :: {path}")
DISPATCH[dtype](text, r)
return r.render_and_exit()
if __name__ == "__main__":
sys.exit(main(sys.argv))
|