agent-skills

Search library/framework documentation via llms.txt (context7.com). Use for API docs, GitHub repository analysis, technical documentation lookup, latest library features.

ALWAYS activate before implementing ANY feature, plan, or fix.

Debug systematically with root cause analysis before fixes. Use for bugs, test failures, unexpected behavior, performance issues, call stack tracing, multi-layer validation, log analysis, CI/CD failures, database diagnostics, system investigation.

ALWAYS activate before fixing ANY bug, error, test failure, CI/CD issue, type error, lint, log error, UI issue, code problem.

Unified knowledge retrieval skill. Build traceable answers from git, MCP, and memory files.

Plan implementations, design architectures, create technical roadmaps with detailed phases.

Apply systematic problem-solving techniques when stuck. Use for complexity spirals, innovation blocks, recurring patterns, assumption constraints, simplification cascades, scale uncertainty.

Apply step-by-step analysis for complex problems with revision capability. Use for multi-step reasoning, hypothesis verification, adaptive planning, problem decomposition, course correction.

Write journal entries analyzing recent changes and session reflections.

Fast codebase scouting using parallel agents. Use for file discovery, task context gathering, searching across directories.

Explore and understand source code using graph_mcp. Always start with semantic_search for fast relevance, then trace call graphs. Default db=neo4j, project_id=hyperdev.

Five expert personas independently analyze proposed changes before implementation to catch architectural, security, performance, and UX issues early. Uses mind_mcp for project context and graph_mcp for code impact analysis. Produces GO/CAUTION/STOP verdict with consensus agreements, conflict resolutions, and risk mitigations. Use before major features, refactors, or risky changes.

Generate comprehensive edge cases and test scenarios by decomposing features across 12 dimensions (user types, input extremes, timing, scale, state, environment, errors, authorization, data integrity, integration, compliance, business logic). Uses mind_mcp for feature requirements context and graph_mcp for code path discovery. Use before implementation, during code review, or when planning test coverage.

STRIDE + OWASP-based security audit with MCP-assisted code analysis and optional iterative auto-fix. Scans code using graph_mcp for structure discovery and mind_mcp for security policy context, then produces severity-ranked findings with fix recommendations. Supports audit-only and audit+fix modes. Use before releases, after sensitive feature additions, or for periodic compliance reviews.

Orchestrate end-to-end deep codebase discovery by combining mind_mcp project knowledge retrieval, graph_mcp semantic and call-graph exploration, and structured synthesis across repo-recon, tech-build-audit, and module-summary-report skills. Use when you need a complete onboarding-quality technical assessment with module mapping, stack/build/platform analysis, critical flows, and prioritized risks.

Build, assess, and improve AI coding agent harness infrastructure across five subsystems — instructions, state, verification, scope, and lifecycle. Use mind_mcp for project context and graph_mcp for code structure discovery when designing harness files. Use when creating AGENTS.md, setting up session continuity, designing verification workflows, or benchmarking agent reliability.

Reconstruct technical documentation from source code by tracing entry points, call flows, IPC links, and domain entities, then produce requirements/use case/detail design artifacts aligned with the project templates. Use when reverse engineering legacy repositories, rebuilding missing specs, or preparing migration-ready docs from implementation-first systems.

Generate structured wiki documentation from source code and project documentation by combining mind_mcp knowledge retrieval with graph_mcp code structure discovery. Produces architecture overview, module reference, API reference, setup guide, and index pages in wiki-compatible markdown. Use when creating onboarding wikis, technical reference sites, or Confluence/GitHub Wiki pages from codebases.

Compute hybrid software bid estimates with best/base/worst person-month ranges and fixed-price/T&M cost ranges using WBS complexity multipliers and capped risk buffers.

Aggregate and normalize evidence for software bidding from mind_mcp, graph_mcp, and trusted internet sources, then emit confidence-ready evidence logs with citations.

Apply proposal lifecycle quality gates for software bidding, covering scope clarity, architecture readiness, estimation confidence, delivery readiness, and production readiness.

Build bidding presentation decks from proposal artifacts using the Presentations skill and Gemini image generation adapter with fallback behavior when API or model checks fail.

Design two software delivery solution options for bidding (baseline and optimized), including architecture direction, tradeoffs, and cost-quality-timeline impacts.

Generate phase-based staffing plans for software bidding using default effort ratios across Discovery, Foundation, Build, Stabilize/UAT, and Go-live/Hypercare.

Orchestrate end-to-end software project bidding packages in hybrid mode (fixed-price + T&M) by coordinating evidence retrieval, solution design, quality gates, estimation, staffing plans, and slide deck generation.

Analyze bugs and evaluate their impact across the codebase by combining mind_mcp context knowledge with graph_mcp dependency tracing. Use when triaging bugs, assessing risk, planning fixes, or estimating regression scope.

Port a single C++ file into Java skeleton form with strict 1-to-1 mapping of package/path/file/class/function/parameter names, plus dependency-aware task breakdown and compatibility requirements with comprehensive security hardening, operational resilience, and evidence-based porting.

Port one C++ function to Java with strict 1-to-1 name preservation, dependency/impact analysis, compatibility-bridge planning, and function-level migration task artifacts with comprehensive security hardening, operational resilience, and evidence-based porting.

Plan migration order for C++ to Java by computing dependency-aware module, file, and function waves; produce an executable worklist with cycle handling, risk scoring, and verification checkpoints with comprehensive security hardening, operational resilience, and evidence-based planning.

Orchestrate end-to-end C++ to Java migration by sequencing pre-porting, file-structure porting, function porting, and legacy guardrails with parity-focused checkpoints and migration deliverables with comprehensive security hardening, operational resilience, and evidence-based orchestration.

Prepare pre-porting analysis artifacts for C++ to Java migration, including compatibility gap inventory, compat-layer design, type mapping, migration roadmap, testing strategy, and risk mitigation outputs in pre-porting-data with comprehensive security hardening, operational resilience, and evidence-based planning.

Unified knowledge retrieval skill that answers with evidence by prioritizing git context, then mind_mcp/graph_mcp, then memory files (memory.md, agent.md, claude/cursor notes). Use when users ask why a change happened, what impacts a function/screen, or need traceable technical context before decisions.

Port and modernize very large legacy C/C++ classes or functions while preserving behavior with explicit guardrails, slice-by-slice migration, and parity tests. Use when files are thousands of lines, functions are hundreds to thousands of lines, logic is stateful or side-effect-heavy, and existing tests are weak or missing.

Synthesize module-level findings into a concise architecture summary using mind_mcp knowledge evidence and graph_mcp semantic/call-graph evidence, highlighting responsibilities, stack, build flow, platform targets, and key risks. Use after repository and tech/build scans when stakeholders need a readable, decision-focused report.

Build a structural understanding of an unfamiliar repository by combining project knowledge-base retrieval from mind_mcp with semantic code exploration from graph_mcp, then produce a module inventory and entry-point map. Use when onboarding to a new codebase, preparing architecture reviews, planning refactors, or creating handover documentation.

Detect core technologies, build systems, CI/CD pipelines, deployment targets, and platform assumptions by combining mind_mcp project knowledge with graph_mcp semantic code evidence. Use when documenting an unknown project stack, preparing migrations, validating onboarding docs, or estimating build and runtime risks.