Stivi7

Stivi7/cyberpunk-context-runners

Cursor rule Cursor

Route project tasks through the runtime-neutral Cyberpunk engineering team.

Stivi7/cyberpunk-context-runners

Agent

These rules apply to every Cyberpunk agent. Role files narrow responsibility but do not override project policy or user instructions.

Stivi7/cyberpunk-context-runners

Agent

Provide the shared engineering contract used by every implementation specialist: bounded scope, repository-native design, test discipline, debugging rigor, and evidence-bearing handoff.

Stivi7/cyberpunk-context-runners

Agent

Act as the backend engineer for services, domain logic, APIs, persistence, migrations, authorization, concurrency, and server-side integrations.

Stivi7/cyberpunk-context-runners

Agent

Turn an approved complex plan into bounded, dependency-aware jobs that can be implemented and integrated safely.

Stivi7/cyberpunk-context-runners

Agent

Independently determine whether a worker result satisfies requirements, project policy, and the evidence bar before integration.

Stivi7/cyberpunk-context-runners

Agent

Implement and review infrastructure, delivery automation, observability, reliability, and operational changes using the project's actual platform tools.

Stivi7/cyberpunk-context-runners

Agent

Pressure-test complex plans before implementation cost accumulates.

Stivi7/cyberpunk-context-runners

Agent

Translate an accepted task brief into the smallest implementable design, interface contract, and verification-aware plan justified by its risk.

Stivi7/cyberpunk-context-runners

Agent

Act as the frontend engineer for user interfaces, interaction behavior, client state, accessibility, responsive layouts, visual consistency, and browser-facing performance.

Stivi7/cyberpunk-context-runners

Agent

Turn one user task into a coherent, verified integration-branch delivery by coordinating the smallest safe adaptive workflow.

Stivi7/cyberpunk-context-runners

Agent

Build and maintain an evidence-based model of the repository so the team follows the project's real stack, commands, conventions, and boundaries.

Stivi7/cyberpunk-context-runners

Skill Claude CodeCodex

Use when backend work affects APIs, domain invariants, persistence, migrations, authorization, concurrency, or compatibility.

Stivi7/cyberpunk-context-runners

Skill Claude CodeCodex

Use when a worker result commit needs independent requirement, correctness, risk, and maintainability review before integration.

Stivi7/cyberpunk-context-runners

Skill Claude CodeCodex

Use when frontend work changes user interactions, visual states, accessibility, responsiveness, or browser-facing behavior.

Stivi7/cyberpunk-context-runners

Skill Claude CodeCodex

Use when a standard or complex change needs explicit architecture, interfaces, sequencing, and verification before implementation.

Stivi7/cyberpunk-context-runners

Skill Claude CodeCodex

Use when work changes infrastructure, delivery automation, permissions, observability, environments, or operational configuration.

Stivi7/cyberpunk-context-runners

Skill Claude CodeCodex

Use when task evidence suggests a durable decision, proven pattern, or reusable lesson should be recorded or updated.

Stivi7/cyberpunk-context-runners

Skill Claude CodeCodex

Use when a complex engineering plan needs adversarial review before implementation begins.

Stivi7/cyberpunk-context-runners

Skill Claude CodeCodex

Use when project stack, commands, conventions, boundaries, or baseline health are unknown or stale.

Stivi7/cyberpunk-context-runners

Skill Claude CodeCodex

Use when implementing an approved work packet inside an isolated worker worktree.

Stivi7/cyberpunk-context-runners

Skill Claude CodeCodex

Use when a test fails, behavior is unexpected, or an implementation attempt does not address the observed problem.

Stivi7/cyberpunk-context-runners

Skill Claude CodeCodex

Use when a new task needs a proportional quick, standard, or complex engineering workflow.

Stivi7/cyberpunk-context-runners

Skill Claude CodeCodex

Use when an approved complex plan must be split into dependency-aware jobs with safe ownership boundaries.