ngpestelos-skills

ActiveJob architecture patterns: fatal vs non-fatal error handling, transaction safety, external API integration, async enqueuing, internal batching, and operation-based idempotency.

Prevents N+1 queries, duplicate joins, and missing eager loading in ActiveRecord. Covers batch preloading, SQL CASE consolidation, 8 N+1 discovery patterns, and join optimization.

Testing patterns for ActiveRecord eager loading: .association(:name).loaded? verification, query count tests, pattern-matching counts, and anti-patterns.

Graceful handling of missing/corrupted foreign key data: .find_by for optional lookups, safe navigation with fallbacks, and callback nil guards with throw(:abort).

Idempotent API endpoints using find_or_create_by! to prevent race conditions and duplicate records. Covers case-insensitive dedup and guard vs idempotent pattern.

Three query performance techniques preventing 10-1000x PostgreSQL degradation: ILIKE index usage, two-phase DISTINCT optimization, and UNION ALL OR-splitting.

Optimizes ActiveRecord transaction boundaries by moving read operations outside transaction blocks to reduce lock time and contention.

Three-agent adversarial validation pattern exploiting LLM sycophancy: Finder (find), Adversary (disprove), Referee (score).

Audit CLAUDE.md files for redundancy, verbosity, stale references, conflicting rules, and optimization opportunities.

Guidance on implementing comprehensive input validation in Rails controller actions before processing to prevent edge case failures.

Prevents double render errors and connection pool pollution in ActiveRecord transaction blocks.

Framework for systematically questioning cultural trends and popular wisdom to reveal strategic insights.

Reduce cyclomatic complexity in Ruby methods via early-return extraction, guard unification, and safe attribute access helpers.

Guidance on what happens when database migrations are terminated mid-execution, how to design safe recoverable migrations, and running long migrati...

Systematic methodology for identifying orphaned files and dead code after refactoring.

Systematic first-principles methodology for debugging complex issues that resist conventional approaches.

Apply Elon Musk's Five-Step Algorithm to systematically optimize any process, system, or workflow through questioning, elimination, optimization, a...

Reusable patterns for stateful Claude Code hooks: priority cascade routing, per-route daily cooldown, CWD-first state fallback, portable date parsi...

Memoize model methods called multiple times per object during view rendering to reduce Ruby CPU and memory allocations.

Fixes the querySelector('form') first-form bug when a Stimulus controller wraps a repeated list where each panel has its own hidden-field form.

Test type selection prioritizing integration tests (89-95% faster than system tests).

Vary response framing by context (learner, engineer, family, researcher) instead of defaulting to consulting/business framing.

Dual-mode investigation framework combining Five Whys (systemic root cause analysis from Toyota/IDEO) and Peeling the Onion (psychological layer ex...

Extract structured data from scanned documents (images, PDFs) into markdown notes, preserving all granular detail.

Quick-reference security checklist for XSS, CSRF, PII logging, console statements, multi-tenant isolation, and state machine security in Rails view...

Systematic methodology for refactoring bloated skills (>500 lines) into focused sub-skills.

Stimulus controller patterns for AJAX initialization, CustomEvent communication, and controller reuse.

Replace AJAX + DOM manipulation with form submission + server redirect.

Systematic 4-phase test planning approach to prevent coverage blind spots.

Guides portable configuration across Claude Code and OpenCode. Covers feature mapping, command mirroring via symlinks and home-manager, architecture layout.

Systematically change configuration defaults and constants while ensuring zero regressions. Covers finding all references, updating code/tests/docs, comprehensive validation.

Guides creating, testing, and deploying Claude Code hooks via nix-darwin. Covers hook input JSON format, exit codes, stdout context injection, transcript JSONL parsing.

Guides updating AI tool templates to use dynamic placeholders for model names instead of hardcoded values in commit attributions.

Systematically analyze uncommitted changes and organize them into logical, atomic commits following best practices.

Gracefully update a git repository from remote when you have unstaged local changes. Stashes changes, pulls, then restores them.

Systematic 5-phase checklist for production deployment: verify completion, prepare commit, create PR, deploy, and monitor.

Role-specific CV variants and cover letters: skills match scoring, reframing, de-AI voice pass, and two-agent quality gate.

Guides Neovim configuration through home-manager programs.neovim module. Covers plugin management, nvim-tree, theme consistency, buffer navigation, and PATH issues.

Guides multi-user nix-darwin configuration with home-manager. Covers mkDarwinConfig, sharedSystemModule, activation scripts, home-manager file conflicts.

Fixes zsh compinit insecure directories warning in nix-darwin. Covers enableGlobalCompInit vs completionInit differences between nix-darwin and home-manager.

Packages pre-built binaries from GitHub releases as Nix derivations for home-manager. Covers stdenv.mkDerivation, fetchurl, SHA256 hash workflow.

Diagnose and resolve disk space issues on macOS with Nix. 5-level cleanup hierarchy from safe Nix GC to caches, with warning signs and Nix-specific symptom table.

Guides deploying nix-template shell.nix and .envrc to directories for per-directory package availability via direnv. Covers home.packages vs devShell distinction.

Guides Node.js and npm version management in Nix flakes. Covers system packages, activation scripts, devShells, version synchronization, pnpm configuration.

Guides fixing build failures in external flake inputs by overriding their derivations with overrideAttrs.

Guides dual-tool configuration for Claude Code and OpenCode. Covers compatibility matrix, single-source commands, skill/agent promotion workflow.

Core patterns for using QMD (Quick Markdown) semantic search. Covers MCP tool usage, search type selection, index rebuild protocol, and troubleshooting.

Uncertainty labeling, evidence hierarchy, and agent output verification. Activates when using absolute language (always, never, prevents, guarantees), making claims about implementation, or consuming agent/subprocess outputs.

Systematic adversarial review of strategic frameworks. Challenges assumptions, identifies correlated failure modes, checks demand evidence.

Eliminates content duplication across Claude Code agent/skill/command layers by establishing clear separation of concerns and reference hierarchies.

Inventory and audit skills for spec compliance and optimization. Two modes: triage (scan all) and deep audit (single skill).

3-question test for auditing custom Claude Code commands: covered by skills?, unique protocol?, actively used? Archive stale/redundant commands to .claude/archives/.

4-step methodology for disruption-proof personal strategy: commoditization-scarcity inversion, pillar architecture with resistance tests, personality alignment filter, uncorrelated wealth streams with binary execution gates.

Pressure-test grand frameworks with 3 tests (falsifiability, independence, predictive specificity) and 4-level Pattern Verification Hierarchy. Distinguishes structural isomorphism from metaphorical resemblance.

Embed abstract strategic frameworks into existing daily/weekly workflow commands as lightweight rituals. 4-step pattern: timescale mapping, daily touchpoint design, weekly audit structure, data flow wiring.

Identify AI-era development priorities via 2-step audit across five irreplaceable human capabilities: Computation, Transformation, Variation, Selection, Attention.

Three-mode first-principles methodology: problem decomposition with Rails/Stimulus debugging patterns, foundation building from observations to laws, and document summarization with length calibration.

Transforms vague existential distress into structured inquiry frameworks with journaling prompts across 7 angles: future, past, counterfactual, relational, existential, synthesis, application.

Self-assessment against the 10 execution habits from 'A Message to Garcia'. Rate, evidence, identify keystone habit, create improvement plan.

Quick-reference toolkit for interrupting rumination patterns with immediate physical and cognitive interventions in 30 seconds or less.

Multi-phase methodology for safely deduplicating large file collections using content-based analysis. Covers analysis, planning, execution with dry-run, verification, and rollback.