Directory
‹ Back to apps

Velkross

Mergeability-first engineering guidance and skills for Cursor.

What’s inside

9 skills
  • Skillagent-memoryUse this skill to read, write, update, search, or organize repo-local Markdown memory notes so project-specific context can be resumed across conversations. Use when the user asks to remember, save, recall, or clean up notes, or when durable project knowledge, restart points, decisions, gotchas, solved problems, or in-progress context should be preserved or checked before related work. Do not use for transient task details, facts that belong in repository docs, or information that is not useful across conversations.
  • Skillagent-skill-designUse when designing, reviewing, or refining Agent Skills, converting workflows or guidance into Skills, deciding whether content belongs in a Skill/tool/docs/policy/memory, or tightening vague, broad, unsafe, or hard-to-invoke Skills so they become reusable, selectable, safe, evaluable, and behavior-changing. Do not use as a substitute for platform-specific packaging, manifest, or authoring-spec rules.
  • Skillclarify-or-proceedDecide whether to ask, assume, investigate, compare options, or proceed when a task has unclear goals, missing constraints, or multiple valid approaches. Use for ambiguous feature requests, design decisions, uncertain implementation paths, and scope-sensitive refactors. Do not use for routine edits, straightforward bug fixes, factual lookups, or tasks with an established local pattern.
  • Skillcomplexity-reviewReview implemented code, diffs, or snippets for unnecessary or unjustified complexity. Use for simplification-focused reviews that look for needless abstractions, unused options, redundant dependencies, avoidable state, duplicated logic, or over-engineering. Avoid for implementation planning, general correctness, security, performance, accessibility, or architecture review.
  • Skilldesign-deep-modulesDesign, reshape, or review module, service, package, API, slice, port, or adapter boundaries using the Deep Modules lens. Use when deciding what a boundary should hide, designing a stable public interface, reducing shallow/pass-through services, separating domain/application policy from infrastructure or framework details, evaluating dependency direction, or planning boundary-focused tests. Avoid for general code review, pure simplification with no boundary/API concern, tiny local bug fixes, visual-only work, or documentation-only tasks that do not define architecture or public contracts.
  • Skillmanage-project-docsMaintain repository docs/ by creating, organizing, auditing, splitting, merging, archiving, or updating project documents. Use when deciding where project knowledge belongs, maintaining docs/README.md as the project-doc router, adding or reviewing frontmatter, or reducing stale or duplicated docs. Do not use merely to read project context. For context discovery, start from docs/README.md directly.
  • Skillpressure-testPressure-test a plan, design, decision, or implementation approach through a focused one-question-at-a-time interview. Use when the user asks to be grilled, wants to validate assumptions, resolve ambiguous tradeoffs, align expectations, or reach shared understanding before implementation. Do not use when the user is asking to implement immediately, when repo context can answer the uncertainty directly, or when deterministic validation or tests are the better next step.
  • SkilltddApply practical test-driven development for adding or changing behavior, fixing bugs with regression tests, refactoring while preserving observable behavior, and code changes that can be driven by examples. Use a small List → Red → Green → Refactor loop, with lightweight variants for UI, integration, database, CLI, file-system, and external-boundary work. Avoid strict TDD for visual-only styling, exploratory spikes, unknown SDK investigation, mechanical migrations, or documentation-only changes unless tests clarify the final behavior.
  • Skillto-issuesConvert PRDs, plans, specs, conversations, backlog ideas, or existing issues into small vertical-slice GitHub/Linear issue drafts for implementation tickets, backlog breakdowns, and agent-ready work items. Use when the requested output is issue drafts or published issues; do not use for broad strategy, prioritization, or architecture work unless it is being turned into issues.