legal-toolkit

Analyze communication patterns from emails, texts, phone records, and chat logs. Builds relationship networks, detects communities, identifies key players, and finds communication gaps or anomalies. Use when: (1) a user provides communication data and asks for analysis, (2) a user says 'analyze these emails', 'map the communication network', 'who are the key players', 'find communication patterns', or 'analyze these phone records', (3) any task involving email analysis, communication pattern mapping, relationship network building, or discovery review, (4) a user wants to understand who communicated with whom, identify clusters or communities, or find gaps in communications around key dates.

Forensic financial analysis of bank statements and transaction records. Traces money flows between entities, detects anomalies (structuring, rapid in-out, unusual timing), and generates interactive visualizations. Use when: (1) a user provides bank statements or transaction data and asks for analysis, (2) a user says 'analyze these transactions', 'trace the money', 'follow the money flow', 'check for suspicious transactions', or 'forensic financial analysis', (3) any task involving bank statement review, transaction tracing, money flow mapping, or financial anomaly detection, (4) a user wants to understand where money went, identify suspicious patterns, or map entity relationships from financial data.

Analyze evidence photos for EXIF metadata, GPS coordinates, timestamps, camera identification, file hashes for integrity verification, and tampering indicators. Generates interactive maps, evidence catalogs, and timelines. Use when: (1) a user has a directory of photos to analyze for legal evidence purposes, (2) a user says 'analyze these photos', 'extract photo metadata', 'check photo EXIF data', 'map photo locations', 'evidence catalog', or 'check for photo tampering', (3) any evidence documentation task involving photographs, (4) a user needs to create an evidence exhibit catalog with metadata and integrity hashes.

Analyze video files frame-by-frame for forensic review with scene detection and multi-pass extraction. Use when: (1) a user provides a video file and wants frame-by-frame analysis, (2) a user says 'analyze this video', 'review body cam footage', 'forensic video review', 'analyze surveillance footage', 'evidence video review', or 'body cam analysis', (3) any forensic or evidentiary task involving video recordings such as body cam, dashcam, surveillance, or cell phone video, (4) a user needs a detailed timeline of events from video footage with key moment flagging and scene change detection.

Audit law firm billing data (LEDES files, Excel invoices, CSV time entries) for compliance issues, block billing, excessive hours, rate violations, and billing anomalies. Use when: (1) a user provides billing files and asks for an audit or review, (2) a user says 'audit these invoices', 'review billing', 'check these LEDES files', 'analyze legal spend', or 'find billing issues', (3) any task involving law firm invoice review, outside counsel billing compliance, or legal spend analysis, (4) a user wants to identify block billing, vague descriptions, excessive hours, duplicate entries, or rate violations.

Build a master case chronology from legal documents. Extracts all dated events from PDFs, DOCX, TXT, and MD files, builds an interactive timeline, detects gaps and date conflicts, and produces a comprehensive chronology spreadsheet. Use when: (1) a user provides a directory of case documents and asks for a chronology or timeline, (2) a user says 'build a chronology', 'create a timeline', 'extract dates from these documents', 'find all events', or 'case timeline', (3) any litigation support task requiring date extraction and event tracking across multiple documents, (4) a user needs gap analysis or date conflict detection across case files.

Calculate court litigation deadlines with FRCP Rule 6 compliance, jurisdiction-aware holiday and business day handling, service method adjustments, and cascading deadline chains. Use when: (1) a user needs to calculate court filing deadlines, response deadlines, or discovery deadlines, (2) a user says 'calculate deadlines', 'when is the response due', 'what are the deadlines for this case', 'FRCP deadlines', or 'court calendar', (3) any litigation scheduling task involving federal or state court rules, (4) a user needs to export deadlines to a calendar (.ics) or generate a deadline report.

Compare two legal documents (PDF, DOCX, TXT) and generate detailed visual diffs with color-coded changes, change heatmaps, and structured change logs. Use when: (1) a user wants to compare two versions of a document, (2) a user says 'compare these documents', 'what changed between these versions', 'show me the differences', 'diff these files', or 'analyze the changes', (3) a user has an original and revised version of a contract, agreement, or policy, (4) a user needs to review edits made to a legal document, (5) a user wants a visual side-by-side comparison.

Index deposition video/audio recordings with transcription, speaker identification, topic indexing, and key moment detection. Use when: (1) a user provides a deposition recording and asks for it to be indexed, transcribed, or analyzed, (2) a user says 'index this deposition', 'transcribe this video', 'find key moments', 'create a depo index', or 'testimony timeline', (3) any litigation support task requiring searchable deposition transcripts with timecodes, (4) a user needs page:line format transcripts, topic indexes, or key moment flagging from recorded testimony.

Extract named entities from legal documents and map relationships between them using NLP. Processes PDF, DOCX, TXT, and MD files or directories of documents. Uses spaCy for named entity recognition to identify people, organizations, dates, monetary amounts, jurisdictions, and legal references, then builds interactive relationship graphs showing how entities connect across documents. Use when: (1) a user provides legal documents and asks to identify entities or map relationships, (2) a user says 'find all entities', 'map relationships', 'who is mentioned in these documents', 'extract names and dates', or 'analyze entity connections', (3) any legal analysis task requiring entity extraction, relationship mapping, or cross-document entity tracking, (4) a user needs to understand which people, organizations, and dates appear across a set of legal documents.

Extract text from scanned PDFs and images using OCR. Uses PaddleOCR (primary, highest accuracy) with pytesseract fallback. Includes image preprocessing (deskewing, contrast enhancement, noise reduction), confidence scoring, and multi-language support. Use when: (1) a user has scanned PDFs or images that need text extraction, (2) a user says 'OCR this document', 'extract text from this scan', 'read this scanned PDF', or 'process this image', (3) document-summarizer reports empty extraction and suggests OCR, (4) a user has a directory of scanned legal documents to batch-process, (5) a user needs to make scanned documents searchable.

Process email archives for e-discovery legal review. Parses .eml, .msg, .mbox files or directories of mixed email files. Extracts metadata, reconstructs threads, identifies duplicates, flags potentially privileged communications, and generates interactive communication network visualizations. Use when: (1) a user provides email files and asks to process them for legal review, (2) a user says 'process these emails', 'analyze this email archive', 'find privileged emails', 'map email communications', or 'prepare emails for review', (3) any e-discovery task involving email parsing, threading, deduplication, or privilege review, (4) a user needs communication network analysis from email archives.

Process raw client intake data into structured legal intake outputs. Handles free-form notes, JSON form data, DOCX documents, and CSV files. Uses NLP to extract client information, classify matter type, prepare conflict check lists, generate document checklists, and calculate statute of limitations deadlines. Use when: (1) a user provides intake notes or client information and asks to process it, (2) a user says 'process this intake', 'structure these client notes', 'prepare conflict check', 'what documents do we need', or 'check statute of limitations', (3) any new client onboarding task requiring structured data extraction from raw notes, (4) a user needs to convert unstructured intake information into organized client profiles and action items.

Generate tracked-changes redline documents from two contract versions (.docx) with risk analysis and change categorization. Use when: (1) a user wants to create a redline of two contract versions, (2) a user says 'redline these contracts', 'generate tracked changes', 'create a redline', 'show me what changed in this contract', or 'compare these contract versions', (3) a user has original and revised .docx contracts and needs Word tracked-changes markup, (4) a user needs risk analysis of contract changes, (5) a user wants to identify material vs administrative changes in a contract revision.

Research SEC EDGAR public filings for companies, extracting financial data, officer/director information, risk factors, and filing histories. Generates research reports with financial trend charts. Use when: (1) a user needs to research a company's public filings or SEC records, (2) a user says 'research this company', 'find SEC filings', 'look up EDGAR records', 'company financials', 'who are the officers', or 'public records search', (3) any corporate due diligence or litigation research task involving SEC data, (4) a user needs financial trends, officer lists, or risk factor analysis from public filings.

Process and summarize large documents (PDF, DOCX, TXT, Markdown) or entire directories of mixed documents. Automatically chunks documents into manageable sections, coordinates a team of agents for parallel summarization, and produces a unified report with executive summary, document structure outline, and section-by-section summaries. Use when: (1) a user provides a file or folder and asks for a summary, overview, analysis, or key takeaways, (2) a user says 'summarize this document', 'summarize these documents', 'give me an executive summary', 'what does this document say', or 'analyze this report', (3) any document processing task where files are too large to read in a single pass, (4) a user points to a directory containing multiple PDFs, DOCX, TXT, or MD files to summarize together.

Transcribe audio or video recordings into professional Word documents with timestamps and speaker labels. Use when a user provides a recording file (.mp3, .wav, .m4a, .mp4, etc.) and wants it transcribed.

Generate a comprehensive DUI trial preparation guide from case files -- reads arrest reports, body cam transcripts, witness statements, lab results, and calibration records, then cross-references against the NHTSA manual to produce a structured trial notebook with chronology, witness profiles, NHTSA compliance analysis, chemical test analysis, inconsistencies, and motion opportunities. Use when: (1) a user provides DUI/DWI case files and asks for trial prep, (2) a user says 'trial prep guide', 'DUI trial notebook', or 'NHTSA analysis', (3) a user provides DUI arrest reports, FST documentation, or BAC results and wants a comprehensive trial-ready analysis.

Analyze criminal defense discovery packages -- police reports, body cam transcripts, witness statements, lab results, calibration records -- and produce a structured defense memo with case snapshot, chronology, NHTSA compliance cross-reference, statutory analysis, inconsistencies, motion opportunities, and recommended next steps. Use when: (1) a user provides discovery documents and wants them analyzed, (2) a user says 'analyze this discovery', 'review this police report', 'cross-reference this arrest report', 'find inconsistencies in these documents', or 'build a defense memo', (3) a user has a DUI/DWI case file to review, (4) a user wants to compare officer reports against NHTSA standards.

Analyze criminal defense case files (police reports, discovery, witness statements, lab results, body cam transcripts) and produce a strategic defense playbook with defense theory, evidence neutralization plan, cross-examination angles, jury considerations, and recommended motions. Use when: (1) a user provides case files and asks for a defense playbook, strategy, or trial prep analysis, (2) a user says 'build a playbook', 'analyze this case', 'defense strategy', or 'trial prep', (3) a user provides charging documents, police reports, or discovery materials and wants strategic analysis.

Draft filing-ready criminal defense motions (suppress, dismiss, exclude, limine, compel, sentencing memo) from case documents and optional templates. Reads case files (PDF, DOCX, TXT), extracts facts, fills templates, produces complete motions with cited facts, IRAC/CRAC arguments, evidence references, and prayer for relief. All case law flagged [VERIFY]. Use when: (1) a user says 'draft a motion', 'write a motion to suppress', 'prepare a motion to dismiss', or similar, (2) a user provides case documents and asks for a motion, (3) a user has a motion template to fill with case facts.

Maps the complete client journey for a law firm from awareness through post-case, identifying pain points, drop-off points, conversion bottlenecks, automation opportunities, and coaching gaps at each stage. Takes firm process documentation or descriptions as input and produces an actionable journey map with prioritized improvement recommendations.

Takes one piece of source content (blog post, article, transcript, or content file) plus a firm voice profile and produces 12+ derivative pieces -- social media posts, email newsletter snippets, Google Business Profile posts, and a short-form video script. Each output is platform-native and matches the firm's voice. Supports audio/video files (chains to transcribe), PDF/DOCX (extracts text), and plain text or URLs.

Analyze attorney caseloads using docket data and complexity weighting. Produces capacity reports per attorney with overload flags, redistribution recommendations, and hiring trigger analysis. Use when a user provides case data, attorney rosters, or staffing details.

Generate adaptive intake call scripts that branch based on charge type, case facts, caller demeanor, and objections. Supports learning from past call recordings (audio/video), transcripts (PDF/DOCX/TXT), or direct practice area input. Use when: (1) a user asks to build, create, or generate an intake script or call guide, (2) a user says 'build intake script', 'create call script', 'intake training materials', or 'call guide for [charge type]', (3) a user provides past call recordings and wants intake scripts derived from them, (4) any task involving intake call design, objection handling scripts, or intake rep training materials.

Build objection response playbooks from intake call transcripts, notes, or audio recordings. Categorizes objections and generates empathetic response scripts for criminal defense intake teams.

Takes charge type, complexity level, jurisdiction, prior record, and aggravating factors as inputs and produces a recommended retainer range with payment plan options, fee justification language, and flags for cases outside historical norms -- always pending attorney review, never a final quote. Use when: (1) a user needs to price a criminal defense case, (2) a user says 'calculate pricing', 'price this case', 'retainer estimate', 'fee quote', 'how much should we charge', or 'payment plan', (3) any intake or case evaluation task involving fee estimation, (4) a user needs payment plan options or fee justification language for a client.

Design a 90-day client communication calendar for law firms based on practice area, case type, estimated timeline, preferred contact frequency, and available channels. Produces scheduled touchpoints, message templates, cadence adjustments, and automation recommendations.

Structure any major firm decision — hiring, case acceptance, pricing, expansion, technology investment, marketing spend — into a clear analysis with options, financial impact, risk assessment, and a recommendation with stated confidence level. Decision support, not decision making.

QA review and anti-hallucination check for legal-toolkit skill outputs. Reads all output files and source materials, verifies factual claims against sources, checks for fabricated citations and unsupported facts, flags inconsistencies, and produces a qa_review.md with severity-rated issues and fixes. Invoked automatically as the final step of every other legal-toolkit skill.

Generates personalized client review request scripts based on case disposition, client relationship quality, and target platforms (Google, Avvo, Yelp). Produces optimal timing recommendations, platform-specific guidance, follow-up sequences, and ensures compliance with bar ethics rules on solicitation of testimonials.

Score intake call transcripts against a customizable rubric for criminal defense firms. Accepts audio/video files (transcribes first), scanned PDFs (OCRs first), regular PDFs, or plain text. Produces weighted scores (1-5) per category with evidence citations, coaching notes with alternative phrases, top strength, top coaching opportunity, objections list, and sign probability.

Transform raw case data exports into a structured performance dashboard -- surfaces KPIs like case volume, resolution time, disposition breakdown, revenue by case type and attorney, intake conversion, and source ROI with period-over-period comparisons and anomaly flags.