Systems That
Do the Work.

CITM Code Review (28-stage pipeline): Standardizes code review across all repositories. Architecture, code quality, logic, branch matrix, function, class, security, performance, and documentation review. Produces measurable findings classified by severity.

Code Intelligence Pipeline: Same 28-stage pipeline applied as a reusable skill. Language-independent: Python, JavaScript, TypeScript, HTML, CSS, JSON, YAML, Bash, SQL.

Engineering Constitution: Governs how engineers and AI agents work. Defines repository structure, documentation requirements, and operational standards before any code is written.

Engineering Workflow: Merges implementer and auditor into one continuous discipline. Every line of code exists because a decision was made and can be defended. Mandatory phase reports.

Global Rules: Session bootstrap protocol. Environment check, dev/prod separation, context ingestion, scope confirmation. Token cost management for large files.

Ponytail System: Anti-bloat engineering persona. Decision ladder: YAGNI, existing implementation, standard library, existing dependency, one-liner, minimum viable system. Zero speculative abstractions.

RCA Engine: Root Cause Analysis system built on the UIOS engine architecture. Hybrid TypeScript + Python stack with multi-agent orchestration.

Energy Fuel Cost Variance RCA: Decomposes energy procurement cost variance into commodity price effect, operational efficiency effect, utilization mix effect, and hedging strategy effect. Distinguishes hedgeable from unhedgeable variance.

Oil & Gas Asset Integrity RCA: Investigates unplanned shutdowns and equipment failures. Addresses systemic causes beyond immediate failure: maintenance interval mismatches, inspection gaps, and digital transformation barriers.

Supply Chain Delivery Risk RCA: Uses structural causal models to distinguish correlated factors from causal drivers. Identifies true bottleneck nodes in multi-country, multi-stakeholder supply chains.

20 reusable skills for business optimization and roadmap planning at scale. Includes competitor analysis, market research, strategic positioning, pipeline architecture porting, open-source license analysis, network-building via free service incentives, wedge pitch design, root cause analyst positioning, post-transaction service design, multi-document synchronization, handover document engineering, and scalability planning.

Applicable to any B2B SaaS, marketplace, or platform business.

Spec-driven development assumes the user already knows the real problem. This protocol handles the harder, earlier case: the user does not yet know the real problem, and the system has to discover it before a specification can be written.

Pipeline: Problem Discovery, Knowledge Construction, Decision Validation, Specification Generation, Architecture, Task Planning, Implementation, Testing, Deployment.

Domain-Agnostic Research Architecture Framework. Maximizes the probability of discovering all decision-relevant information while minimizing search bias and blind spots.

Core principle: Research quality is a function of Search Space, Search Strategy, Source Coverage, and Validation. Fits as a foundational layer beneath CITM, PURF, and UIOS.

Innovation Proposal Development Framework (IPDF): 11-phase pipeline that transforms an existing research repository into an investment-grade competition proposal. Every claim is traceable to repository evidence. Produces application summary, concept note, and supporting evidence package.

Competition Response Workflow: Ties brain systems, rapid response kit, and IPDF into a single end-to-end pipeline. Fast path (5-7 hours) for standard challenges, full path (days to weeks) for complex ones.

8-section frontend code review protocol. Run after every build phase. No phase advances until all sections pass.

Covers: layout and alignment, accessibility, performance, SEO and meta, security, cross-browser, responsive design, and code quality. Synthesizes rules from Front-End Checklist, Dify Accessibility, React Doctor, and Firefox frontend best practices.

Transforms existing PPTX templates into polished, data-driven presentations using only authoritative project data from the local repository.

Audits every template, reverse-engineers the design system, identifies placeholders and dummy data, searches the repository for correct source data, and builds a presentation narrative around available evidence.

Security Gate (STBP): Cautious senior engineer posture. Defends against secret leakage, destructive operations, unauthorized exposure, supply-chain risk, and prompt injection. Least privilege by default.

Validation Gate (PIVP): Context lock, dead code sweep, automated debug pass, full test suite, strictest linter. No phase advances until this gate passes.

Agent Handoff Protocol: Structured handoff packet for session transitions. Status first, decisions made, open items, do-not-touch list, open questions, evidence.

Stack Migration Protocol: Behavior-preserving translation, not rewrite from memory. Full functional inventory, characterization tests, concept mapping table.