| name | t1k:contract-review |
| description | Adversarial contract review and negotiation playbook. Use for reviewing term sheets, partnership agreements, distribution/licensing contracts, NDAs BEFORE signing. Modes: (1) one-sided tilt analysis — surface clauses that disadvantage you; (2) cross-party synthesis — reconcile your review with counterparty's markup; (3) round-based balance review — multi-pass polish of your own draft toward a desired posture (neutral / slightly-company-favorable / slightly-partner-favorable); (4) term rationale doc — why-it-exists / what-each-side-gets / tension-axis for each major clause; (5) Schedule A genre scope drafting; (6) round-2 cover email structure (post-rebalance send). Complements t1k:contract (which handles rendering/linting). Use this skill FIRST (review + draft) before running t1k:contract (render + beautify). |
| keywords | ["contract review","term sheet review","negotiation review","redline","adversarial contract","balance review","tilt analysis","cross-party synthesis","term rationale","schedule a","non-compete scope","cover email","post-rebalance","round 2 send","partnership agreement review","distribution agreement review","licensing agreement review","nda review","counter-offer","clause-by-clause"] |
| argument-hint | tilt <file.md> | synthesize <our-review.md> <their-review.md> | balance <file.md> <posture> | rationale <file.md> | schedule-a <doc.md> | cover-email <doc.md> |
| effort | medium |
| version | 1.94.2 |
| origin | theonekit-core |
| repository | The1Studio/theonekit-core |
| module | t1k-extended |
| protected | true |
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-contract:
- This skill (
contract-review) is about what the contract says — clause analysis, tilt detection, redlining, cover email drafting.
t1k-contract is about how the contract looks — pandoc rendering, lint gotchas, CSS, tabularization.- Typical workflow: review with
contract-review → edit the Markdown → t1k-contract beautify to produce the final .docx / .pdf.
Modes
| 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-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
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)
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
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
- Received a draft from counterparty.
- Run
tilt from your side. Output: deal-killer / structural / polish list.
- Counterparty replies with their markup.
- Run
synthesize over your tilt + their feedback. Output: overlap / divergence / blind-spot lists.
- Optionally run
pattern-recognize for strategic-level pattern classification.
- Decide your target posture (neutral / slightly-you / slightly-them).
- Draft your revised version.
- Run
balance in rounds R1 → R2 → R3. Write delta log each round.
- 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.
- Before sending, document rationale.
- Run
rationale to produce {scope}/internal/term-rationale-EN.md.
- If Schedule A is referenced but missing, draft it.
- Run
schedule-a using the 4-section template.
- Draft the cover email.
- Run
cover-email with the round-2 structure; include internal fallback posture.
- Render and send.
- Hand off to
t1k-contract beautify to produce .docx / .pdf.
- Use the sending checklist in the cover email draft to verify attachments.
Gotchas
- Tilt review is not redlining. Tilt produces a severity-ranked list; redlining produces a specific line-by-line edit. Do tilt first; redline in a separate pass.
- Synthesize only after you have both reviews. Don't run synthesize on just one side — you'll bias toward whoever you read first.
- Balance posture must be explicit. If you can't articulate "slightly-company-favorable" vs "neutral," you don't know what you're optimizing for.
- Rationale doc is internal, not for sending. If counterparty asks for rationale, produce a counterparty-facing summary — not the full doc.
- Schedule A must be updatable, not one-shot. Without A.3 (annual update), the schedule ossifies and becomes a liability when Company changes direction.
- Cover email fallback section is INTERNAL. Never paste the fallback posture into the actual email. Remove before sending.
- Scenario divergences are features, not bugs. Don't flatten them in balance unless the divergence was accidental.
- Cross-reference staleness after
balance — after renumbering sections, grep for § cross-refs and verify each still points to the right section. balance R3 should include this sweep.
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-contract instead.
- Drafting from scratch with no counterparty yet. Use a template library first; contract-review is for review + negotiation, not greenfield drafting.
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 — 12-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
Created from patterns observed across the following counterparties:
- Alfa Partnership (advisor equity / sales partner) — originating patterns: Schedule A genre scope, round-2 cover email, balance review discipline.
- BagelCode Pre-Publishing Agreement — deal-killer pattern: Bagelcode-side IP assignment on Phase 1 funds (from
BagelCode/funding-agreement/).
- Neptune — deal-killer pattern: IP-hostage risk elevated to critical (from
Neptune/ session).
- JM Game distribution — bilingual-contract pattern: edits must update both Chinese and English.
This skill is designed to be counterparty-agnostic; the modes above apply to any commercial negotiation with meaningful leverage on both sides.
