Skills › Knowledge Work Plugins › contract-review › gotchas.md

gotchas.md

105 lines · small-business/skills/contract-review/reference/gotchas.md at 2574e29

Open on GitHub

1# Gotchas

3Edge cases in contract analysis. Good / Bad pairs.

5---

Gotchas

Edge cases in contract analysis. Good / Bad pairs.

Raw lines 1 to 6

7## Gotcha: Flagging standard boilerplate as a red flag

9**Why it matters:** Over-flagging trains the user to ignore the summary. If everything is a red flag, nothing is.

Gotcha: Flagging standard boilerplate as a red flag

Why it matters: Over-flagging trains the user to ignore the summary. If everything is a red flag, nothing is.

Raw lines 7 to 10

11### ✗ Bad

13```

14Contract has a standard mutual NDA confidentiality clause, 2-year duration.

15Claude: "🔴 Red flag: Confidentiality clause imposes obligations on both parties."

16```

18Standard mutual NDA is market-norm. Flagging it as a red flag destroys signal-to-noise ratio.

✗ Bad

Contract has a standard mutual NDA confidentiality clause, 2-year duration.
Claude: "🔴 Red flag: Confidentiality clause imposes obligations on both parties."

Standard mutual NDA is market-norm. Flagging it as a red flag destroys signal-to-noise ratio.

Raw lines 11 to 19

20### ✓ Good

22```

23Contract has a standard mutual NDA confidentiality clause, 2-year duration.

24Claude: [skips; does not flag]

25```

27Reserve red flags for genuinely non-standard terms. If a clause is boilerplate and fair, omit it.

29---

✓ Good

Contract has a standard mutual NDA confidentiality clause, 2-year duration.
Claude: [skips; does not flag]

Reserve red flags for genuinely non-standard terms. If a clause is boilerplate and fair, omit it.

Raw lines 20 to 30

31## Gotcha: Missing clauses aren't missing from the analysis

33**Why it matters:** For SMBs, the absence of standard protections (liability cap, change order process, consequential damages waiver) is often more dangerous than an unfavorable clause — the default legal position fills the gap, usually in the counterparty's favor.

Gotcha: Missing clauses aren't missing from the analysis

Why it matters: For SMBs, the absence of standard protections (liability cap, change order process, consequential damages waiver) is often more dangerous than an unfavorable clause — the default legal position fills the gap, usually in the counterparty's favor.

Raw lines 31 to 34

35### ✗ Bad

37```

38Contract has no liability cap.

39Claude: [no mention; only analyzes clauses that are present]

40```

42User signs thinking liability is limited. It isn't.

✗ Bad

Contract has no liability cap.
Claude: [no mention; only analyzes clauses that are present]

User signs thinking liability is limited. It isn't.

Raw lines 35 to 43

44### ✓ Good

46```

47Contract has no liability cap.

48Claude: "🔴 Missing: No liability cap. This contract contains no limitation-of-liability

49clause. Under default law, your exposure is uncapped. Standard practice is to cap at

50fees paid in the prior 12 months. Suggest adding: 'Each party's total liability shall

51not exceed the fees paid or payable in the 12 months preceding the claim.'"

52```

54Treat absent-but-standard clauses as red flags. Explicitly label them "Missing."

56---

✓ Good

Contract has no liability cap.
Claude: "🔴 Missing: No liability cap. This contract contains no limitation-of-liability
clause. Under default law, your exposure is uncapped. Standard practice is to cap at
fees paid in the prior 12 months. Suggest adding: 'Each party's total liability shall
not exceed the fees paid or payable in the 12 months preceding the claim.'"

Treat absent-but-standard clauses as red flags. Explicitly label them "Missing."

Raw lines 44 to 57

58## Gotcha: Large PDFs truncated mid-analysis

60**Why it matters:** Contracts over 20–30 pages often have the most dangerous terms in exhibits, schedules, or "Order Forms" at the back. If the PDF is read only up to page 15, key terms are silently missed.

Gotcha: Large PDFs truncated mid-analysis

Why it matters: Contracts over 20–30 pages often have the most dangerous terms in exhibits, schedules, or "Order Forms" at the back. If the PDF is read only up to page 15, key terms are silently missed.

Raw lines 58 to 61

62### ✗ Bad

64```

65Skill reads pages 1–10 of a 40-page MSA. Schedules A–D (starting page 28) contain

66the IP assignment and liability cap. Skill produces a "clean" summary.

67```

✗ Bad

Skill reads pages 1–10 of a 40-page MSA. Schedules A–D (starting page 28) contain
the IP assignment and liability cap. Skill produces a "clean" summary.

Raw lines 62 to 68

69### ✓ Good

71```

72Read the PDF in chunks: pages 1–10, 11–20, 21–40. Analyze all chunks before

73producing the summary. If a section heading mentions "Schedule," "Exhibit," or

74"Appendix," flag it explicitly and ensure it was read.

75```

77Always read the full document. Use the `pages` parameter to chunk large PDFs.

79---

✓ Good

Read the PDF in chunks: pages 1–10, 11–20, 21–40. Analyze all chunks before
producing the summary. If a section heading mentions "Schedule," "Exhibit," or
"Appendix," flag it explicitly and ensure it was read.

Always read the full document. Use the pages parameter to chunk large PDFs.

Raw lines 69 to 80

81## Gotcha: Counterparty template vs. negotiated draft

83**Why it matters:** Recommendations should match the negotiating context. Pushing back hard on a startup's first-draft agreement is different from pushing back on a Fortune 500 procurement template — the latter is often non-negotiable on 80% of its terms.

Gotcha: Counterparty template vs. negotiated draft

Why it matters: Recommendations should match the negotiating context. Pushing back hard on a startup's first-draft agreement is different from pushing back on a Fortune 500 procurement template — the latter is often non-negotiable on 80% of its terms.

Raw lines 81 to 84

85### ✗ Bad

87```

88Fortune 500 vendor agreement with standard procurement terms.

89Claude: "🔴 This Net-60 payment term should be renegotiated to Net-30.

90 Their legal team will likely accept the change."

91```

93Unrealistic recommendation wastes the user's political capital.

✗ Bad

Fortune 500 vendor agreement with standard procurement terms.
Claude: "🔴 This Net-60 payment term should be renegotiated to Net-30.
         Their legal team will likely accept the change."

Unrealistic recommendation wastes the user's political capital.

Raw lines 85 to 94

95### ✓ Good

97```

98Claude: "🟡 Net-60 payment terms. This is longer than typical (Net-30 is standard),

99 but common in large enterprise procurement templates. Worth asking for

100 Net-45 as a compromise — they may accept it, especially for recurring work.

101 Note: if this is a Fortune 500 template, payment terms are often non-negotiable

102 in the first engagement."

103```

104

105Match the recommendation to the power dynamic. Label it yellow, not red. Acknowledge the reality.

✓ Good

Claude: "🟡 Net-60 payment terms. This is longer than typical (Net-30 is standard),
         but common in large enterprise procurement templates. Worth asking for
         Net-45 as a compromise — they may accept it, especially for recurring work.
         Note: if this is a Fortune 500 template, payment terms are often non-negotiable
         in the first engagement."

Match the recommendation to the power dynamic. Label it yellow, not red. Acknowledge the reality.

Raw lines 95 to 105