leninkhaidem
- 17 skills
- 0 followers
- 16 hours ago last updated
- ▌ Architecture Diagrams · leninkhaidem bundleUse when creating professional architecture diagrams, cloud infrastructure visuals, network topologies, Kubernetes cluster diagrams, or microservices architecture diagrams as PNG/SVG images using Python Diagrams library with real provider icons (AWS, Azure, GCP, K8s, OnPrem, Generic)
- ▌ Audit · leninkhaidem bundleFinal read-only planned-feature completion audit. Use when the user asks to audit, verify implementation, check completion, validate the build, or confirm the feature matches the accepted plan. Do not use as ordinary code review or to repair files inline.
- ▌ Release · leninkhaidem bundlePrepares or publishes completed work with version bumps, tags, release notes, and cleanup. Use when asked to prepare, publish, version, tag, ship, or finish a release. Do not use for development, code review, audit, or unrelated cleanup.
- ▌ Testing · leninkhaidem bundleEstablish, document, and apply project-specific testing workflows. Use for testing strategy, workflow setup/revision, test authoring/alteration, or execution. Do not use for bug fixes, feature implementation, code review, audit, or release.
- ▌ Code Doc · leninkhaidem bundleGenerates comprehensive documentation for any codebase via hybrid analysis (native extractors + LLM agents). Triggers on phrases like "document this codebase", "generate documentation", "create docs for this repo", "write documentation", "document the code", "generate codebase docs", "create architecture docs", "write a developer guide", "document this project".
- ▌ Worktree · leninkhaidem bundleGit worktree strategy for branch-isolated development. Use for planned-feature package work, isolated bugfixes, hotfixes, spikes, feature-branch management, or worktree cleanup. Do not use for direct implementation without an approved worktree action.
- ▌ Implement · leninkhaidem bundleExecutes reviewed Slice-first planned-feature packages for approved changes. Use when asked to implement, execute, build, or continue an approved planned-feature package workflow. Do not use for plan authoring, plan review, ordinary PR review, audit, or dashboard status.
- ▌ Review Code · leninkhaidem bundleReviews code changes with bounded multi-agent analysis. Use for PRs, local diffs, or planned-feature final integrated review. Do not use for final planned-feature audit or PR/local repairs unless the selected mode explicitly permits fixes.
- ▌ Review Plan · leninkhaidem bundleValidates Slice-first planned-feature artifacts for initial approval or focused same-requirement re-review during implementation. Use to review, validate, or approve a plan. Do not use to perform implementation, code review, audit, dashboard status, or ordinary PR review.
- ▌ Perspectives · leninkhaidemThis skill should be used when the user asks to "explore approaches", "get multiple perspectives", "divergent analysis", "think about this from different angles", "architecture decision", "devil's advocate", or faces a complex design decision, architectural problem, or technical challenge. Triggers on phrases like "perspectives on", "multiple angles", "explore options", "compare approaches". Also appropriate when architecting a new product or facing ambiguous cross-cutting problems.
- ▌ Conceptualize · leninkhaidem bundleExplore a product, architecture, or research idea through rigorous one-question-at-a-time discovery until shared understanding is reached. Use when the user asks to conceptualize, shape an idea before planning, stress-test direction, collect research/context, or prepare mandatory planning Slices. Do not use when the user wants implementation, code review, audit, or task-dashboard status.
- ▌ Readme Polish · leninkhaidem bundleAuthor/polish a single repository README and optional GitHub/GHE repo metadata when asked to create/update README, improve repo polish, About/description, website, topics, or social preview. Near miss: whole-codebase or multi-file docs route to code-doc.
- ▌ Empirical Spike · leninkhaidemProduces bounded evidence for one unresolved empirical behavior. Use directly or when planning, plan review, or implementation depends on observed behavior after static evidence is insufficient. Do not use to perform implementation, code/plan review, or routine testing.
- ▌ Skill Authoring · leninkhaidem bundleCreates or revises compact, self-contained skills that mid-tier agents can execute reliably. Use when adding, rewriting, or structurally reviewing a skill. Do not use to perform its domain task.
- ▌ Walk Me Through · leninkhaidemInteractive, evidence-grounded teaching of features, PRs, releases, and code changes for practical ownership. Use only when explicitly invoked by the user. Not for implementation, code review, audits, or bulk documentation.
- ▌ Diagnose And Fix · leninkhaidem bundleDiagnoses defects evidence-first, obtains human-readable fix authorization, and routes repairs. Use for bugs, failing tests, regressions, troubleshooting, or "fix this". Do not use for planned features, ordinary review, containment, docs, releases, or speculative cleanup.
- ▌ Implementation Plan · leninkhaidem bundleCreates or repairs Slice-first planned-feature artifacts for approved changes. Use for initial planning, same-requirement plan repair during auto-resolve implementation, package breakdowns, or task artifacts. Do not use to perform coding, code review, audit, or status.