omeeragtoprak
UserVerification-driven operating protocol for AI coding agents - Explore/Plan/Implement/Verify/Deliver with adversarial review and a hard verify gate. Claude Code plugin + portable AGENTS.md (Codex, Cursor, Gemini CLI).
Categories
Indexed Skills (9)
deliver
Delivery discipline — evidence-backed summary, atomic commit etiquette, and updating the project's persistent memory (§P) with durable knowledge. Use when finishing a task, preparing commits or a PR, or when the user asks to wrap up, summarize, or commit the work.
explore
Read-only deep exploration and skeptical gap analysis before any code is written. Use at the start of every non-trivial task, when onboarding to an unfamiliar codebase or module, or when the user asks "how does X work here", "analyze before changing", or requests an As-Is/To-Be gap analysis.
implement
Production-grade implementation standards — typing, naming, error handling, concurrency, dependencies, configuration, and atomic scope. Use while writing or modifying any code, and consult when reviewing code for standards violations.
orchestrate
Lead-engineer delegation — decompose work into parallelizable contracts, brief subagents without leaking your conclusions, isolate contexts, and integrate results by verifying behavior rather than trusting reports. Use when a task spans many files/subsystems, when independent verification is needed, or when fan-out across subagents would beat serial work.
plan
Structured planning with a generate→critique→refine brainstorming loop, a short tech spec, and approval gates for risky changes. Use after exploration on any multi-file, uncertain, or unfamiliar change, or when the user asks for options, alternatives, a design, or an implementation plan.
protocol
Run the full Agentic Engineering Protocol (Explore → Plan → Implement → Verify → Deliver) on a task. Use for any non-trivial engineering task — new features, bug fixes, refactors, migrations — or when the user says "use the protocol", "AEP", or asks for disciplined/production-grade execution.
research
Deep, source-triangulated web research inside engineering work — decompose the topic into sub-branches, search from multiple angles, require independent sources, check recency and versions, and label everything unverified. Use before choosing a library/API/architecture, when working with fast-moving or unfamiliar technology, for security advisories, or whenever a decision rests on a claim not proven in this repository.
standards
The AEP standards reference — security baseline (OWASP-aligned), performance and database discipline, and testing standards. Consult when touching anything web-facing, security-sensitive, performance-critical, or database-related, and during code review.
verify
Ruthless verification — closed feedback loop until green, regression tests, adversarial review in a fresh context, gap-closure audit, and an evidence block. Use after any implementation, before declaring any task done, or when the user asks "is this actually done/correct/tested".
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.