t1k:business:contract-review
| Field | Value |
|---|---|
| Module | base |
| Version | 0.12.1 |
| Effort | medium |
| Tools | — |
Keywords: adversarial contract, balance review, clause-by-clause, contract review, cover email, cross-party synthesis, distribution agreement review, licensing agreement review, nda review, negotiation review, non-compete scope, partnership agreement review, post-rebalance, redline, round 2 send, schedule a, term rationale, term sheet review, tilt analysis
How to invoke
Section titled “How to invoke”/t1k:business:contract-reviewtilt <file.md> | synthesize <our-review.md> <their-review.md> | balance <file.md> <posture> | rationale <file.md> | schedule-a <doc.md> | cover-email <doc.md>Contract Review — Adversarial Negotiation Playbook
Section titled “Contract Review — Adversarial Negotiation Playbook”A structured review and drafting playbook for commercial contracts where both parties have real leverage — term sheets, partnership agreements, distribution agreements, licensing agreements, advisor equity agreements, NDAs with substantive commercial exposure.
Scope boundary vs t1k-business-contract:
- This skill (
contract-review) is about what the contract says — clause analysis, tilt detection, redlining, cover email drafting. t1k-business-contractis about how the contract looks — pandoc rendering, lint gotchas, CSS, tabularization.- Typical workflow: review with
contract-review→ edit the Markdown →t1k-business-contract beautifyto produce the final.docx/.pdf.
| Mode | What it does | When to use | Detail |
|---|---|---|---|
tilt | One-sided review: read the contract from your side only; surface every clause that disadvantages you, ranked by severity (deal-killer / structural / polish). | Received a draft from the counterparty; haven’t drafted your own markup yet. | references/mode-1-tilt.md |
synthesize | Merge your tilt review with the counterparty’s own markup/counter-feedback; identify overlap (both sides flagged), divergence (only one flagged), and blind spots (neither flagged but material). | Counterparty has replied with their own feedback; you need to see the whole negotiation surface. | references/mode-2-synthesize.md |
balance | Multi-round polish of a draft toward a target posture (neutral / slightly-you-favorable / slightly-them-favorable). Rounds are structured: R1 deal-killer fixes, R2 structural moderation, R3 polish, R4 edge-case adversarial, R5 cross-doc consistency. | Preparing your own draft to send; posture is a deliberate choice. | references/mode-3-balance.md + references/second-layer-defects.md (R4/R5) |
rationale | For each major clause, produce a why-it-exists / what-each-side-gets / tension-axis analysis. Output doc lives at {scope}/internal/term-rationale-EN.md. | Preparing to walk the counterparty through your reasoning (sales-y), or teaching your own team why each clause is shaped as it is. | references/mode-4-rationale.md |
schedule-a | Draft a non-compete genre-scope appendix: explicit in-scope genres + explicit out-of-scope carve-outs + annual update mechanism + interpretation rules. | Non-compete clause references a Schedule A but none is attached. | references/mode-5-schedule-a.md + references/schedule-a-template.md |
cover-email | Draft a round-2 cover email (post-rebalance send): what-moved-your-way / what-held-and-why / why-leading-with-X / attachments / internal fallback posture section. | Sending a revised draft back to the counterparty after their counter. | references/mode-6-cover-email.md + references/cover-email-round2-template.md |
pattern-recognize | Read counterparty’s markup/feedback and classify the negotiation pattern they’re using (1 of 20 named patterns). Cross-links to t1k-business-negotiation for full pattern catalog + counter-tactics. | After receiving any counterparty markup; complement to tilt mode for strategic-level (not just clause-level) understanding. | references/mode-7-pattern-recognize.md |
Each mode is documented in detail in its corresponding reference file. Read the relevant file before running the mode for the first time in a session.
Cross-mode patterns
Section titled “Cross-mode patterns”Canonical deal-killer clauses
Section titled “Canonical deal-killer clauses”A 10-clause recurring pattern across many contract reviews. Always scan for these in tilt; always fix in balance R1. Full list with rationale and replacement language: references/canonical-deal-killers.md.
Quick reference:
- Voluntary exit forfeits vested equity
- Non-compete >18 months, worldwide, no genre carve-out
- Clawback with no fair-cause exceptions
- Unilateral termination, <30-day notice, no acceleration
- Gross margin / cost-of-sales as a vesting gate for a sales partner
- Pool B (exit) with minimum-valuation cliff
- IP assignment without carve-out for Partner’s pre-existing methodologies
- “Material breach” undefined in for-cause termination
- Exit / valuation ladder without strict
≥and<boundary inequalities - Anti-non-compete jurisdictions without firm-level fallback
Second-layer defects (R4/R5 adversarial pre-send pass)
Section titled “Second-layer defects (R4/R5 adversarial pre-send pass)”After R1-R3 of balance fix obvious deal-killers, a second layer of defects remains — clauses that look fine in isolation but fail an adversarial reading. Categories: contradictions between clauses; asymmetric safety valves; closed exception lists missing realistic causes; undefined terms that can be weaponized; boundary mechanics breaking at the edge; loopholes in time-based mechanics; non-cash transaction handling; cross-document numerical drift.
Always run R4 (edge-case) and R5 (consistency + polish) before sending. Typical find rate: 5–10 must-fix + 15–20 should-fix per contract. Full scan list with examples: references/second-layer-defects.md.
Intentional scenario divergence — when to preserve
Section titled “Intentional scenario divergence — when to preserve”If a contract has multiple scenarios, resist the urge to flatten them during balance. Preserve divergences that are structurally motivated (different KPI shape, different per-unit equity weight, different revenue recognition). Document the divergence in the rationale doc and balance log. Detail: references/second-layer-defects.md.
Workflow — typical session
Section titled “Workflow — typical session”- Received a draft from counterparty.
- Run
tiltfrom your side. Output: deal-killer / structural / polish list.
- Run
- Counterparty replies with their markup.
- Run
synthesizeover your tilt + their feedback. Output: overlap / divergence / blind-spot lists. - Optionally run
pattern-recognizefor strategic-level pattern classification.
- Run
- Decide your target posture (neutral / slightly-you / slightly-them).
- Draft your revised version.
- Run
balancein rounds R1 → R2 → R3. Write delta log each round.
- Run
- Adversarial pass before sending.
- Run R4 (edge-case stress test) and R5 (cross-doc consistency + polish). Use the scan list in
references/second-layer-defects.md. Do not skip this — it catches the defects that survive the first 3 rounds. - For parallel execution, 5 independent sub-agents (one per round: Company-exposure / Partner-fairness / edge-cases / consistency / polish) can run concurrently; aggregate into a consolidated action plan.
- Run R4 (edge-case stress test) and R5 (cross-doc consistency + polish). Use the scan list in
- Before sending, document rationale.
- Run
rationaleto produce{scope}/internal/term-rationale-EN.md.
- Run
- If Schedule A is referenced but missing, draft it.
- Run
schedule-ausing the 4-section template.
- Run
- Draft the cover email.
- Run
cover-emailwith the round-2 structure; include internal fallback posture.
- Run
- Render and send.
- Hand off to
t1k-business-contract beautifyto produce.docx/.pdf. - Use the sending checklist in the cover email draft to verify attachments.
- Hand off to
When NOT to use this skill
Section titled “When NOT to use this skill”- Template contracts with no real negotiation surface (e.g., cookie-cutter SaaS EULA, standard NDA). Use a lawyer-provided template; don’t over-engineer.
- Contracts in regulated domains with statutory form requirements (e.g., Vietnamese labor contracts, consumer credit contracts, real estate). The framework may not map cleanly; consult local counsel.
- Pure rendering/formatting work — use
t1k-business-contractinstead. - Drafting from scratch with no counterparty yet. Use a template library first; contract-review is for review + negotiation, not greenfield drafting.
References
Section titled “References”Mode-detail files (read before first use of each mode):
references/mode-1-tilt.md— severity buckets, output format, 12-clause checklistreferences/mode-2-synthesize.md— 3-list structure, output format, disciplinereferences/mode-3-balance.md— postures, round structure, log template, scenario divergencereferences/mode-4-rationale.md— 4-part analysis, output template, divergence tablesreferences/mode-5-schedule-a.md— 4-section structure, drafting rules, common mistakesreferences/mode-6-cover-email.md— section outline, internal-only sections, tone disciplinereferences/mode-7-pattern-recognize.md— workflow, classification heuristic, worked example
Cross-mode patterns:
references/canonical-deal-killers.md— 10-item canonical list with rationalereferences/second-layer-defects.md— R4/R5 adversarial pre-send scan list
Standalone templates (drop into a contract project’s templates/ or internal/ folder):
references/deal-killer-checklist.md— 10-clause recurring tilt checklist with example languagereferences/schedule-a-template.md— 4-section templatereferences/cover-email-round2-template.md— round-2 cover email templatereferences/balance-review-log-template.md— delta log template
Provenance
Section titled “Provenance”Patterns synthesized from 4 real counterparty reviews (Alfa Partnership, BagelCode, Neptune, JM Game) — full per-deal pattern attribution in references/provenance.md.