aicontracts

AICONTRACTS — AI Contract Review

Capability overview

Code: AICONTRACTS Full name: AI Contract Review Description: AI contract review — clause extraction, risk scoring, confidence-gated exports, comparison to market standard.

---

Skills index

#	Skill	Command	Purpose
1	Clause Extraction	/cps-aicontracts:clause-extract	Extract named clauses (indemnity, limitation of liability, IP, termination, etc.) with confidence scores.
2	Risk Scoring	/cps-aicontracts:risk-score	Score extracted clauses against playbook standards; flag deviations.
3	Redline Suggestions	/cps-aicontracts:redline	Propose redlines to bring deviations into playbook compliance.
4	Contract Comparison	/cps-aicontracts:compare	Compare two contracts clause-by-clause; produce delta report.

---

Skill 1: Clause Extraction

Command: /cps-aicontracts:clause-extract

Purpose

Extract named clauses (indemnity, limitation of liability, IP, termination, etc.) with confidence scores.

Inputs required

client:
  name: "Client Name"
  industry: "Industry"
context:
  scope: "in-scope description"
  constraints: ["constraint 1", "constraint 2"]
references:
  - "Prior deliverable / document name"
  - "External benchmark / source"

Methodology

1. Frame. Translate client request into a structured problem statement and hypotheses.
2. Gather. Pull data from engagement brain, client documents, and external sources.
3. Analyze. Apply the Clause Extraction method using domain frameworks (see Frameworks section of the agent).
4. Synthesize. Produce primary deliverable with recommendations, evidence, and risks.
5. Validate. Run /cps:verify-quality and pass through Every extracted clause has confidence score; <0.85 is flagged for human review.

Output shape

A CPS-branded Clause Extraction deliverable in 05_Deliverables_Final/. Pyramid-Principle structured, sourced, and reviewed.

Quality checklist

• Every extracted clause has confidence score; <0.85 is flagged for human review
• Risk scores reference an explicit playbook clause / threshold
• Redlines are tracked against the original; no silent edits

---

Skill 2: Risk Scoring

Command: /cps-aicontracts:risk-score

Purpose

Score extracted clauses against playbook standards; flag deviations.

Inputs required

client:
  name: "Client Name"
  industry: "Industry"
context:
  scope: "in-scope description"
  constraints: ["constraint 1", "constraint 2"]
references:
  - "Prior deliverable / document name"
  - "External benchmark / source"

Methodology

1. Frame. Translate client request into a structured problem statement and hypotheses.
2. Gather. Pull data from engagement brain, client documents, and external sources.
3. Analyze. Apply the Risk Scoring method using domain frameworks (see Frameworks section of the agent).
4. Synthesize. Produce primary deliverable with recommendations, evidence, and risks.
5. Validate. Run /cps:verify-quality and pass through Every extracted clause has confidence score; <0.85 is flagged for human review.

Output shape

A CPS-branded Risk Scoring deliverable in 05_Deliverables_Final/. Pyramid-Principle structured, sourced, and reviewed.

Quality checklist

• Every extracted clause has confidence score; <0.85 is flagged for human review
• Risk scores reference an explicit playbook clause / threshold
• Redlines are tracked against the original; no silent edits

---

Skill 3: Redline Suggestions

Command: /cps-aicontracts:redline

Purpose

Propose redlines to bring deviations into playbook compliance.

Inputs required

client:
  name: "Client Name"
  industry: "Industry"
context:
  scope: "in-scope description"
  constraints: ["constraint 1", "constraint 2"]
references:
  - "Prior deliverable / document name"
  - "External benchmark / source"

Methodology

1. Frame. Translate client request into a structured problem statement and hypotheses.
2. Gather. Pull data from engagement brain, client documents, and external sources.
3. Analyze. Apply the Redline Suggestions method using domain frameworks (see Frameworks section of the agent).
4. Synthesize. Produce primary deliverable with recommendations, evidence, and risks.
5. Validate. Run /cps:verify-quality and pass through Every extracted clause has confidence score; <0.85 is flagged for human review.

Output shape

A CPS-branded Redline Suggestions deliverable in 05_Deliverables_Final/. Pyramid-Principle structured, sourced, and reviewed.

Quality checklist

• Every extracted clause has confidence score; <0.85 is flagged for human review
• Risk scores reference an explicit playbook clause / threshold
• Redlines are tracked against the original; no silent edits

---

Skill 4: Contract Comparison

Command: /cps-aicontracts:compare

Purpose

Compare two contracts clause-by-clause; produce delta report.

Inputs required

client:
  name: "Client Name"
  industry: "Industry"
context:
  scope: "in-scope description"
  constraints: ["constraint 1", "constraint 2"]
references:
  - "Prior deliverable / document name"
  - "External benchmark / source"

Methodology

1. Frame. Translate client request into a structured problem statement and hypotheses.
2. Gather. Pull data from engagement brain, client documents, and external sources.
3. Analyze. Apply the Contract Comparison method using domain frameworks (see Frameworks section of the agent).
4. Synthesize. Produce primary deliverable with recommendations, evidence, and risks.
5. Validate. Run /cps:verify-quality and pass through Every extracted clause has confidence score; <0.85 is flagged for human review.

Output shape

A CPS-branded Contract Comparison deliverable in 05_Deliverables_Final/. Pyramid-Principle structured, sourced, and reviewed.

Quality checklist

• Every extracted clause has confidence score; <0.85 is flagged for human review
• Risk scores reference an explicit playbook clause / threshold
• Redlines are tracked against the original; no silent edits

---

Cross-skill workflows

Discovery → /cps-aicontracts:clause-extract → /cps-aicontracts:risk-score → Recommendations

Integration with other CPS plugins

Plugin	Integration point
cps-ai-dd	Contract review during legal due diligence
cps-ai-regs	Regulatory clauses (e.g., GDPR DPAs)

Templates & artifacts

All deliverables use standard CPS branding via:

• /doc-gen for document generation
• assets/cps-branding.json for styling
• scripts/cps-document-generator.js for automation

Skill	Primary artifact	Format
/cps-aicontracts:clause-extract	Clause Extraction Report	DOCX/PDF
/cps-aicontracts:risk-score	Risk Scoring Report	DOCX/PDF
/cps-aicontracts:redline	Redline Suggestions Report	DOCX/PDF
/cps-aicontracts:compare	Contract Comparison Report	DOCX/PDF