Self Improving Legal

Self-Improving Legal Skill

Log legal learnings, compliance findings, contract issues, and regulatory changes to markdown files for continuous improvement. Captures clause risks, compliance gaps, precedent shifts, contract deviations, regulatory changes, and litigation exposure. Important learnings get promoted to clause libraries, compliance checklists, regulatory trackers, contract playbooks, or risk registers.

CRITICAL: Privilege & Confidentiality Protection

NEVER log privileged attorney-client communications, specific case strategy, confidential settlement terms, or litigation work product in any learning or issue entry. Always abstract findings to process-level lessons. Use generalized descriptions:

When in doubt, omit the specific detail and log only the process improvement lesson. Privilege waiver through careless documentation is irreversible.

First-Use Initialisation

Before logging anything, ensure the .learnings/ directory and files exist in the project or workspace root. If any are missing, create them:

mkdir -p .learnings
[ -f .learnings/LEARNINGS.md ] || printf "# Legal Learnings\n\nClause risks, compliance gaps, precedent shifts, regulatory changes, and legal operations insights.\n\n**Categories**: clause_risk | compliance_gap | precedent_shift | contract_deviation | regulatory_change | litigation_exposure\n**Areas**: contracts | compliance | litigation | intellectual_property | privacy | regulatory | corporate_governance\n\n---\n" > .learnings/LEARNINGS.md
[ -f .learnings/LEGAL_ISSUES.md ] || printf "# Legal Issues Log\n\nContract disputes, compliance failures, regulatory impacts, and litigation matters.\n\n---\n" > .learnings/LEGAL_ISSUES.md
[ -f .learnings/FEATURE_REQUESTS.md ] || printf "# Legal Operations Feature Requests\n\nLegal technology, automation, and process improvement requests.\n\n---\n" > .learnings/FEATURE_REQUESTS.md

Never overwrite existing files. This is a no-op if .learnings/ is already initialised.

Do not log privileged communications, specific case strategy, or confidential settlement terms. Abstract all entries to process-level lessons.

If you want automatic reminders, use the opt-in hook workflow described in Hook Integration.

Quick Reference

OpenClaw Setup (Recommended)

OpenClaw is the primary platform for this skill. It uses workspace-based prompt injection with automatic skill loading.

Installation

Via ClawdHub (recommended):

clawdhub install self-improving-legal

Manual:

git clone https://github.com/jose-compu/self-improving-legal.git ~/.openclaw/skills/self-improving-legal

Workspace Structure

OpenClaw injects these files into every session:

~/.openclaw/workspace/
├── AGENTS.md          # Legal workflows, contract review delegation
├── SOUL.md            # Legal principles, privilege protection, confidentiality
├── TOOLS.md           # Legal tool capabilities, CLM configs, e-discovery
├── MEMORY.md          # Long-term memory (main session only)
├── memory/            # Daily memory files
│   └── YYYY-MM-DD.md
└── .learnings/        # This skill's log files
    ├── LEARNINGS.md
    ├── LEGAL_ISSUES.md
    └── FEATURE_REQUESTS.md

Create Learning Files

mkdir -p ~/.openclaw/workspace/.learnings

Then create the log files (or copy from assets/):

• LEARNINGS.md — clause risks, compliance gaps, regulatory changes, precedent shifts
• LEGAL_ISSUES.md — contract disputes, compliance failures, litigation matters
• FEATURE_REQUESTS.md — legal technology and automation requests

Promotion Targets

When legal learnings prove broadly applicable, promote them:

Optional: Enable Hook

For automatic reminders at session start:

cp -r hooks/openclaw ~/.openclaw/hooks/self-improving-legal
openclaw hooks enable self-improving-legal

See references/openclaw-integration.md for complete details.

Generic Setup (Other Agents)

For Claude Code, Codex, Copilot, or other agents, create .learnings/ in the project or workspace root:

mkdir -p .learnings

Create the files inline using the headers shown above. Avoid reading templates from the current repo or workspace unless you explicitly trust that path.

Add reference to agent files

Add to AGENTS.md, CLAUDE.md, or .github/copilot-instructions.md:

Legal Self-Improvement Workflow

When legal issues or findings occur:

1. Log to .learnings/LEGAL_ISSUES.md, LEARNINGS.md, or FEATURE_REQUESTS.md
2. NEVER log privileged communications, case strategy, or settlement terms
3. Review and promote broadly applicable findings to:
   • Clause libraries — reusable clause language and fallback positions
   • Compliance checklists — regulatory requirement tracking
   • Contract playbooks — negotiation patterns and escalation guides
   • Risk registers — recurring risk categories and mitigation plans

Logging Format

Learning Entry [LRN-YYYYMMDD-XXX]

Append to .learnings/LEARNINGS.md:

## [LRN-YYYYMMDD-XXX] category

**Logged**: ISO-8601 timestamp
**Priority**: low | medium | high | critical
**Status**: pending
**Area**: contracts | compliance | litigation | intellectual_property | privacy | regulatory | corporate_governance

### Summary
One-line description of the legal insight or finding

### Details
Full context: what clause, regulation, or precedent was involved, why it matters,
what the correct approach or fallback position is. NEVER include privileged
attorney-client communications or confidential settlement terms.

### Recommended Action
Specific clause revision, compliance step, filing, or process change

### Metadata
- Source: contract_review | regulatory_update | case_law | audit | negotiation | litigation
- Jurisdiction: US-Federal | US-CA | EU | UK | other
- Related Files: path/to/contract.docx, path/to/policy.md
- Tags: tag1, tag2
- Regulation: GDPR | CCPA | SOX | HIPAA | EU-AI-Act (if applicable)
- See Also: LRN-20260110-001 (if related to existing entry)
- Pattern-Key: clause.indemnity_cap | comply.ccpa_deletion (optional)
- Recurrence-Count: 1 (optional)
- First-Seen: 2026-01-15 (optional)
- Last-Seen: 2026-01-15 (optional)

---

Categories for learnings:

Legal Issue Entry [LEG-YYYYMMDD-XXX]

Append to .learnings/LEGAL_ISSUES.md:

## [LEG-YYYYMMDD-XXX] issue_type

**Logged**: ISO-8601 timestamp
**Priority**: critical | high | medium | low
**Status**: pending
**Area**: contracts | compliance | litigation | intellectual_property | privacy | regulatory | corporate_governance
**Severity**: critical | high | medium | low

### Summary
Brief description of the legal issue

### Issue Details
Context of the legal issue: what happened, what contract or regulation is involved,
what the exposure or risk is. Abstract to process-level lessons — do NOT include
privileged communications, specific case strategy, or confidential terms.

### Impact Assessment
- Affected contracts/matters
- Financial exposure estimate (range, not exact if confidential)
- Regulatory risk (enforcement action, fine range, reputational)
- Deadline or statute of limitations implications

### Recommended Action
- Immediate steps to mitigate
- Escalation path (in-house counsel, outside counsel, board)
- Long-term process improvement

### Timeline
- **Identified**: ISO-8601
- **Escalated**: ISO-8601 (if applicable)
- **Resolved**: ISO-8601 (if applicable)

### Metadata
- Trigger: contract_redline | regulatory_filing | case_law | audit_finding | infringement_notice | dsar
- Jurisdiction: US-Federal | US-CA | EU | UK | other
- Counterparty: [generalized role, e.g., "SaaS vendor", "enterprise customer"]
- Related Files: path/to/document
- See Also: LEG-20260110-001 (if related)

---

Feature Request Entry [FEAT-YYYYMMDD-XXX]

Append to .learnings/FEATURE_REQUESTS.md:

## [FEAT-YYYYMMDD-XXX] capability_name

**Logged**: ISO-8601 timestamp
**Priority**: medium
**Status**: pending
**Area**: contracts | compliance | litigation | intellectual_property | privacy | regulatory | corporate_governance

### Requested Capability
What legal operations tool, automation, or capability is needed

### Business Justification
Why it's needed — what risk it mitigates, what workflow it improves, what compliance
requirement it satisfies

### Complexity Estimate
simple | medium | complex

### Suggested Implementation
How this could be built: CLM integration, regulatory API, compliance dashboard, clause library

### Metadata
- Frequency: first_time | recurring
- Related Features: existing_tool_or_feature
- Compliance: GDPR | CCPA | SOX | HIPAA (if applicable)

---

ID Generation

Format: TYPE-YYYYMMDD-XXX

• TYPE: LRN (learning), LEG (legal issue), FEAT (feature request)
• YYYYMMDD: Current date
• XXX: Sequential number or random 3 chars (e.g., 001, A7B)

Examples: LRN-20260412-001, LEG-20260412-A3F, FEAT-20260412-002

Resolving Entries

When an issue is resolved, update the entry:

1. Change **Status**: pending → **Status**: resolved
2. Add resolution block after Metadata:

### Resolution
- **Resolved**: 2026-01-16T09:00:00Z
- **Action Taken**: Brief description of resolution (abstracted, no privileged details)
- **Clause/Policy Updated**: Reference to updated clause library or compliance checklist
- **Verified By**: legal review | compliance audit | outside counsel

Other status values:

• in_progress — Actively being negotiated, reviewed, or remediated
• wont_fix — Risk accepted with documented justification and approver
• promoted — Elevated to clause library, compliance checklist, or contract playbook
• promoted_to_skill — Extracted as a reusable skill

Detection Triggers

Automatically log when you encounter:

Contract Review Signals (→ learning or legal issue):

• Contract redline rejected 3+ times on the same clause
• Unlimited liability or uncapped indemnity accepted
• Non-standard payment terms without VP approval
• Auto-renewal clause without termination notice window
• Assignment clause missing change-of-control protections
• Most-favored-nation clause absent from strategic vendor agreement

Regulatory Signals (→ learning or legal issue):

• New regulation published affecting the organization's industry
• Regulatory filing deadline within 30 days
• Amendment to existing regulation (GDPR, CCPA, SOX, HIPAA)
• Enforcement action against peer company in same sector
• Cross-border data transfer mechanism invalidated

Litigation Signals (→ legal issue):

• IP infringement notice or cease-and-desist received
• Data subject access request (DSAR) received
• Subpoena or litigation hold notice
• Non-compete or non-solicitation dispute
• Breach of contract claim or demand letter

Compliance Signals (→ learning):

• GDPR/CCPA audit finding
• SOX control deficiency
• Internal audit exception
• Third-party vendor compliance gap
• Data retention policy violation
• Missing consent mechanism for data processing

Precedent Signals (→ learning):

• Court ruling changes interpretation of standard clause
• Regulatory guidance document issued
• Industry association publishes new model terms
• Arbitration award sets new benchmark in relevant area

Priority Guidelines

Area Tags

Use to filter learnings by legal domain:

Promoting to Permanent Legal Standards

When a learning is broadly applicable (not a one-off matter), promote it to permanent legal operations standards.

When to Promote

• Clause risk recurs across multiple contract negotiations
• Compliance gap applies to the entire regulatory program
• Precedent shift affects standard contract positions
• Contract negotiation pattern should be institutionalized
• Regulatory change requires ongoing monitoring

Promotion Targets

How to Promote

1. Distill the learning into a concise clause, checklist item, or playbook entry
2. Add to appropriate target (clause library, compliance checklist, playbook)
3. Update original entry:
   • Change **Status**: pending → **Status**: promoted
   • Add **Promoted**: clause_library (or compliance_checklist, contract_playbook, risk_register)