The skills-meta pack provides foundational and cross-cutting capabilities: task exploration, verified research, problem solving, design thinking, feature pipelines, knowledge harvesting, and security review. Consult the public package source for the current inventory.
Adaptive task exploration and clarification skill for transforming vague requests into actionable specifications. Triggers on any task that needs clarification before execution.
goap-research-ed25519
Advanced GOAP research system with Ed25519 cryptographic verification for anti-hallucination protection. Combines Goal-Oriented Action Planning with cryptographic signatures to ensure source authenticity, claim verification, and citation chain integrity. Use for high-stakes research requiring verifiable facts, competitive intelligence, or legal/medical contexts.
adversarial-verifier
Adversarial verification of claims and outputs — stress-tests findings, hunts for counter-evidence, and flags unsupported assertions before they are trusted.
Elite problem-solving system combining first-principles thinking, game theory, root cause analysis, and TRIZ inventive principles. Use when facing complex challenges with unclear solutions, strategic decisions with multiple stakeholders, or problems requiring breakthrough rather than incremental solutions.
structured-reasoning
Structured, step-by-step reasoning scaffolds for complex multi-part problems — decomposes a question, reasons through each part, and assembles a defensible conclusion.
reflection-loop
Standalone critique-revise cycle (DRAFT → CRITIQUE → IDENTIFY → REVISE → VERIFY → DECIDE) with domain-specific criteria for code, text, architecture, and research. Max 3 rounds with anti-patterns against cosmetic-only or infinite revision.
context-window-management
Strategies for operating under context pressure — priority-based pruning, checkpoint-restore, summarization, sliding window, and delegation. Activates around ~60% capacity and picks the strategy that fits the task.
Human-centered product development orchestrator implementing Stanford d.school’s 5-phase Design Thinking with a 6th Validate phase. Chains explore, GOAP research, JTBD/CJM/VSM analysis, HADI hypothesis cycles, iterative prototyping, and usability testing.
feature-adr
11-step feature development pipeline with Complexity Router (S/M/L/XL), ADR-driven architecture decisions, DDD modeling, QCSD quality swarm, SPARC-GOAP planning, brutal-honesty review, and multi-agent fleet QE. Use for any non-trivial feature implementation.
Multi-agent knowledge harvesting system that extracts reusable patterns, commands, rules, templates, and snippets from any project directory. Uses 5 parallel extractor agents, 7-category classification, and 8 blocking quality gates with auto-placement.
understand-anything-bridge
Bridges the Understand-Anything Claude Code plugin with DZ Harness skills. Reads .understand-anything/knowledge-graph.json and provides architectural context to feature-adr, design-thinking, qe-test-generation, and AgentDB.
capture-adr
Lightweight ADR capture mid-conversation — detects architectural decisions made implicitly during coding, records them as minimal MADR 4.0 entries in docs/decisions/, and tags them needs-oversight for human confirmation.
skill-crystallizer
Auto-creates new skills from execution traces, combines existing skills into composites, and repairs broken skills. Three modes: CAPTURED (trace → skill), DERIVED (combine skills), FIX (detect + repair).
Whole-codebase deep audit — five phases (parallel recon → adversarial self-challenge → P0–P3 ranking → your approval → scoped fixes). Surfaces real bugs/security/perf/dead-code/gaps as a tight numbered list. Triggers on /audit, “audit the codebase”, “health check”, “what should we fix next”.
skill-advisor
Recommends which skills / presets / npx toolkits from the arsenal fit a task — with pipeline order, install commands, and honest gaps. Triggers on “which skill should I use”, “recommend skills for…”, /skill-advisor.