237 Skills for Claude Code

Audit components and pages for WCAG 2.1 AA/AAA accessibility compliance. Use when checking contrast ratios, ARIA labels, keyboard navigation, touch targets, or preparing for accessibility certificatio

Transform words into letter-acrostic algorithms and sentences into word-step algorithms. Word Mode: each letter → procedural step. Sentence Mode: each content word → procedural step. Passive trigger r

Analyze algorithms generated by the algorithmify output style. Scores sequential coherence, domain fit, verb selection, step density, and prohibited-word compliance. Word Mode and Sentence Mode suppor

Run prospective root strength analysis — the ouverse of the conservation law, tracing existence forward into what it enables.

Detect, block, and learn from code antipatterns using an immune system model. Use when repeated bad patterns appear in generated code, or when building self-correcting guardrails. Do NOT use for one-o

Codon Rule — track triplet arc position (DREAM/REALIZE/CRITIQUE). Sub-skill of rhythm. DNA reads in codons of 3; sessions read in arcs of 3. Triggers on '/arc', 'arc position', 'which phase', 'session

Query autopsy_records in brain.db for session outcome trends and root cause patterns. Use when authoring PDP prompts, reviewing directive progress, or analyzing which session patterns correlate with o

Apply Haeckel's biological principles (biogenetic law, germinal layers) to Claude Code architecture via cross-domain transfer. Use when analyzing session bootstrap patterns or architectural stratifica

Design hook-skill molecular structures using chemistry bond metaphors. Use when planning compound hooks, defining bond types between skills and hooks, or computing molecular stability of hook chains.

Apply Boundary-Parameterized Interaction Theory (BPIT) to analyze any entity-system interaction by encoding the perturbation pattern once and projecting through multiple evaluation boundaries. Use whe

Apply boundary definition, classification, and conservation principles extracted from 27 peer-reviewed DNA structure papers (1981-2025) and verified computationally.

Run comprehensive brain integrity audit covering implicit JSON schemas, dual-store divergence, and orphaned data. Use when brain MCP tools fail, after recovery, or when data seems corrupted. Do NOT us

Quick reference for NexCore Brain working memory — sessions, artifacts, code tracking, and implicit learning. Use when saving artifacts, managing sessions, tracking code changes, or using implicit kno

Brain system SOP with 5-phase BRAIN DAG, dual-store architecture, and V2 sync pipeline. Use when executing brain sync operations, managing artifact lifecycle, or troubleshooting dual-store divergence.

Detect and resolve dual-store divergence between brain.db and implicit JSON files. Use when implicit MCP tools return stale data, brain_implicit_get fails, or JSON/SQLite stores disagree. Do NOT use f

Foundry bridge B1→B2: transform design specs into compilable source code. Use when a DesignSpec artifact needs translation to Rust source files. Do NOT use standalone — only within a foundry pipeline.

Foundry bridge A1→A2: transform raw metrics into pattern signatures. Use when numeric measurements need grouping into emergent structural patterns. Do NOT use standalone — only within a foundry analys

Foundry bridge B*→A1: extract measurable metrics from builder source artifacts. Use when code artifacts need quantification for analysis. Do NOT use standalone — only within a foundry pipeline. Trigge

Foundry bridge A3→B1: translate analysis findings into design constraints, closing the foundry feedback loop. Use when intelligence reports need conversion to actionable DesignSpec constraints. Do NOT

Foundry bridge A2→A3: derive causal relationships from pattern data. Use when pattern signatures need causal inference to produce intelligence reports. Do NOT use standalone — only within a foundry an

Foundry bridge B2→B3: validate source code against design contracts and test requirements. Use when compiled code needs verification against its DesignSpec. Do NOT use standalone — only within a found

Manage corrective and preventive actions (CAPA) through the full quality lifecycle. Use when detecting nonconformances, investigating root causes, planning corrective actions, or verifying effectivene

Measure the AlgoVigilance capability compound using the 15-primitive edge ontology. Use when user says '/compound', '/capability-graph', '/kg measure', 'measure capabilities', 'compound health', 'ρ-rate

Enforce the Cartouche newtype pattern for wrapping raw primitives in Rust. Use when creating domain IDs, enforcing type-safe boundaries, or encapsulating values with validation. Do NOT use for typesta

Execute the 8-stage Cognitive Evolution Pipeline for systematic knowledge building with T1/T2/T3 primitive extraction and validation. Use when building cross-domain knowledge, extracting primitives fr

Design, evaluate, and evolve checklists using cognitive architecture from aviation, medicine, nuclear, and SRE domains. Use when creating operational checklists, diagnosing checklist failures, or appl

Reference for chemistry primitives mapped to PV and hook architecture via cross-domain transfer. Use when applying Arrhenius, Michaelis-Menten, Gibbs, Beer-Lambert, or Henderson-Hasselbalch equations

Run the C-L-E-A-N vocabulary program for code hygiene and dead code removal. Use when cleaning dead code, linting, removing empty files, or archiving stale artifacts. Do NOT use for code quality metri

Quick reference for nexcore-cloud 35-type taxonomy and MCP tools. Use when analyzing cloud primitives, comparing types, tracing transfers, or querying infrastructure topology. Do NOT use for GCloud CL

Run FDA auditor-inspired code audits using the 8-step SOP-9001 protocol with Form 483 observations and CAPA generation. Use when performing systematic code inspections or preparing quality evidence. D

Compiled safety SOP enforcing code safety across Rust, bash, TypeScript, and git from brain corrections and learned antibodies. Use when checking safety rules before commits, reviewing unsafe patterns

Apply cognitive bias models (Dunning-Kruger, Loss Aversion, Imposter Syndrome, Sunk Cost, and 6 more) to diagnose judgment distortions. Use when analyzing why a session went wrong, reviewing decision-

Generate context-aware communication guidance showing the user how to interact with Claude Code effectively. Self-bootstrapping — Claude invokes this to teach users its own interface. Use when onboard

Route to communication sub-skills from Schwartzberg's "Get to the Point!" covering point definition, delivery, discipline, amplification, and scenarios. Use when improving presentations, emails, or pe

Optimize text density using the Compendious Score (Cs = I/E x C x R) and BLUFF method. Use when compressing verbose output, measuring information density, or tightening written communication. Do NOT u

Reference for the Apollo-NexCore 9-layer deterministic PV pipeline with conservation laws L1-L4 and typed interface contracts. Use when composing signal-to-response chains, checking conservation invar

Apply compound effects mental models (Lindy, Pareto, Compound Interest, Network Effects, and 4 more) to system analysis. Use when evaluating exponential growth, edge effects, or systemic leverage in a

Detect when a dominant primitive feeds itself through different boundaries, creating compound growth (e.g., ν² where internal use frequency drives external return frequency). Identifies self-reinforci

Execute the autonomous capability compounding loop — learning, synthesis, distillation — to grow validated skills and beliefs. Use when checking autonomy score, reviewing beliefs, or running the full

Search for concepts across naming variations by auto-expanding terms into OR-patterns covering case, abbreviations, and section markers. Use when a simple Grep misses results due to naming inconsisten

Reference for NexCore Config type-safe configuration consolidation. Use when consolidating scattered JSON/INI configs, validating MCP server paths, or migrating to unified TOML format. Do NOT use for

Strategic framework for hub config dominance using Gilded Age monopoly archetypes, Shapiro-Varian standards war theory, and gated execution sequence. 5 clusters: Titan Playbook, Game Theory Frames, St

Sustained-effort reflex ensuring every session cycle spends tokens on work that advances the conservation strategy (∂-building, ∅-filling, or ∃-generating). Concepts that remain concepts produce no ex

Scaffold a conservation law directory structure (void/, state/, boundary/, existence/product/) with INDEX.md template. Use when creating a new knowledge project where content naturally separates into

This skill should be used when the user asks to "create a construct map", "map a learning progression", "define competency levels", "build an assessment dimension", "show expertise continuum", "map co

Execute the 8-stage Constructive Epistemology Pipeline integrating SECI, Bloom, and Double-Loop Learning for systematic knowledge building. Use when converting tacit knowledge to explicit, building st

Detect false convergence in multi-item fix campaigns where clearing easy items masks remaining hard items. Applies boundary-weighted distance instead of naive item counts. Use mid-campaign when progre

Run a self-terminating 5-agent review loop (Architect, Sentinel, Optimizer, Advocate, Validator) that converges when no agent makes edits. Use for deep pre-merge review, architecture audit, or securit

Run the C-R-A-F-T vocabulary program for code quality metrics — coverage, readability, architecture, fitness, and testing. Use when measuring code quality, checking coverage, or reviewing test health.

Scaffold nexcore crates following the gold standard pattern (nexcore-cloud). Use when creating a new crate, auditing crate structure, or scaffolding domain primitives with Cargo.toml, lib.rs, and test

Evaluate external crate dependencies before adding them to the workspace. Decision matrix with MSRV, license, security, and overlap gates plus weighted scoring. Use when considering a new dependency.

Validate software deliverables using the Clinical Trial Validation Paradigm (5-phase: Preclinical through Surveillance) with Five Problems Protocol and Reality Gradient scoring. Use when assessing pro

Orchestrate E2E testing for Studio using Cypress MCP tools. Use when writing end-to-end tests, debugging Cypress runs, or verifying Studio UI flows. Do NOT use for unit tests — use test-protocol inste

Run data transformation pipelines — validate input, transform format, persist to database, and cleanup. Use when processing ETL operations or input validation workflows. Do NOT use for webhook event p

Validate components against the AlgoVigilance Golden Ratio design system with STEM-grounded thresholds for color, spacing, typography, and animation. Use when reviewing UI code or auditing design consis

Route to the correct domain-specific -dev reference skill based on technology or context. Use when unsure which dev skill to load or asking 'which reference for X'. Do NOT use for skill ecosystem ques

Run the D-I-S-T-I-L-L vocabulary program to synthesize knowledge packs into nested skills. Use when converting compiled knowledge into actionable skills or running the distillation pipeline. Do NOT us

Enforce the Djed typestate pattern for compile-time state machine transitions in Rust. Use when designing lifecycle types, state transition guards, or making invalid states unrepresentable. Do NOT use

Translate concepts between domains bidirectionally using T1/T2/T3 primitive mappings with confidence scores. Use when mapping ideas from one domain to another or asking 'what is the equivalent of X in

Run the E-D-I-C-T vocabulary program for gate-locked directive execution across 5 phases. Use when executing architecture directives or managing phase-gated development. Do NOT use for general project

Enforce ? error propagation over unwrap()/expect() in Rust code. PostToolUse gate for Edit/Write on .rs files.

Per-event input/output schemas for all 17 Claude Code hook events. Use when looking up event field names, matcher values, or input/output structure for a specific hook event. Do NOT use for hook proto

Stop Codon — execute session exhale NOW. Writes the 3-line verdict, updates autopsy_records with model-authored verdict, releases the ribosome. The exhale IS the deliverable, not an afterthought. UAA.

Run Kolb's Experiential Learning Cycle (CE→RO→AC→AE) on any target to diagnose learning imbalances and extract principles. Use when reflecting on session outcomes, diagnosing repeated mistakes, or str

Decide between hooks, skills, subagents, and MCP for extending Claude Code capabilities. Use when choosing the right extensibility mechanism, configuring permissions, or designing automation pipelines

Apply fail-safe design patterns from mechanical engineering (Williams 1909) where loss of power/signal/connection defaults to safe state. Use when designing error recovery, graceful degradation, or sa

Orchestrate cross-session feedback via HARVEST→TRIAGE→ENRICH→INJECT pipeline. Use when harvesting session reflections, triaging quality trends, or injecting improvements into future sessions. Do NOT u

Operate the three-node flywheel bridge — execute `/flywheel` (full health), `/flywheel status` (layout), `/flywheel promote NAME` (tier advance), `/flywheel snapshot` (bus events), `/flywheel health`

Orchestrate autonomous Rust feature development via START FORGE protocol with miner→codifier→validator agent loop. Use when building complete Rust features end-to-end with subagent orchestration. Do N

Provide Rust pattern knowledge for FORGE Phase 2 code generation. Use only within a FORGE pipeline when the codifier agent needs pattern reference. Do NOT use standalone — invoke forge instead. Trigge

Provide domain knowledge for FORGE Phase 1 primitive extraction. Use only within a FORGE pipeline when the miner agent extracts primitives. Do NOT use standalone — invoke forge instead. Triggers on 'f

Provide quality gate knowledge for FORGE Phase 3 validation. Use only within a FORGE pipeline when the validator agent runs quality checks. Do NOT use standalone — invoke forge instead. Triggers on 'f

Run the dual-pipeline foundry assembly line with 6 typed stations (B1-B3 builder + A1-A3 analyst) and 8 bridge contracts. Use when building features that need both construction and analysis pipelines.

Mine NCBI GEO for gene expression data via Entrez esearch/efetch. Use when searching for drug treatment expression datasets or differential expression data to support mechanism hypotheses. Do NOT use

Address review and issue comments on the current branch's GitHub PR using gh CLI. Use when PR comments need responses or code changes. Do NOT use for CI failures — use gh-fix-ci instead. Triggers on '

Debug and fix failing GitHub Actions CI checks by pulling logs, summarizing failures, and implementing fixes after approval. Use when PR checks fail on GitHub Actions. Do NOT use for PR review comment

Run the G-L-E-A-N vocabulary program for data source intelligence and insight mining. Use when analyzing external data sources, deriving insights from databases/APIs/websites, or mapping data to goals

Apply the Golf Edict governance framework — validate decisions against G1 (state acceptance), G2 (boundary respect), G3 (proportional judgment), and ρ (self-governance). Routes to axiom checks, penalt

Run the G-U-A-R-D vocabulary program for build quality gates and safety governance. Use when running pre-release checks, validating builds, or enforcing safety gates. Do NOT use for code quality metri

Helix Turn — structural review every ~10 sessions. DNA twists every 10.5 bp; without turns, 2m of DNA can't compact into a 6um nucleus. Compacts artifacts, checks topology, measures drift, scans loops

Amplify hook infrastructure with subagent integration, tier-based enforcement, and molecular bonding architecture. Use when designing compound hooks, building hook pipelines, or integrating hooks with

Diagnose and repair hook infrastructure failures including state file corruption, stale guardian data, and broken scripts. Use when hooks fail, produce errors, or state files are corrupted. Do NOT use

Detect and fix hook wiring gaps where scripts exist on disk but settings.json points to stubs, placeholders, or wrong paths. Use when hooks seem registered but do not fire. Do NOT use for hook logic e

Complete hook system reference covering bash scripts, protocol, event schemas, guardrails, and reflex system. Use when creating hooks, debugging hook behavior, choosing events, or designing hook chain

Enforce the Horus parse-don't-validate pattern for Rust API boundaries. Raw inputs are validated at the boundary and converted to trusted types that interior functions accept unconditionally. Use when

Automate AlgoVigilance GCE infrastructure — deployment, backups, and monitoring. Use when deploying to GCE, managing backups, or checking infrastructure status. Do NOT use for local development — this i

Analyze code integrity using adversarial mutation transformations inspired by FDA auditing. Use when generating mutation tests, scoring structural integrity, or detecting anti-pattern surfaces. Do NOT

Spawn a 5-agent parallel intelligence team for deep multi-perspective analysis. Use when conducting competitive intelligence, capability extraction, or strategic analysis requiring parallel agent pers

Apply railway signalling interlock patterns (Williams 1909) to make invalid states unrepresentable at compile time. Use when designing mutual exclusion constraints, preventing conflicting configuratio

Search past sessions, prompts, tool calls, errors, and propositions via the introspection engine (FTS5 + structured transcript extraction). Use when recalling past work, finding how a problem was solv

Extended Jupyter operations that compose MCP primitives and Jupyter REST API into higher-level commands: Run All, Restart & Run All, Export, Change Kernel, Settings, and Extension management. Use when

Compile scattered documentation into versioned Brain artifacts, resolve to immutable snapshots, and dispose of redundant source files. Use when consolidating knowledge, creating immutable snapshots, o

Execute the two-step exile verb (COMPOST then ARCHIVE) on stale or false knowledge entries across implicit stores. Scans beliefs, corrections, trust, and patterns for staleness using 5 failure modes,

Load knowledge files from ~/.claude/knowledge/ into the MCP knowledge engine by reading index.json, ingesting files as fragments, and compiling domain packs. Bridges durable flat files to the queryabl

Run a color-encoded human-AI knowledge relay session using 20Q decomposition/composition with ChromaTerm visual encoding. Use when user says '/relay', '/knowledge-relay', '/kr', 'let's relay on [topic

Run the L-E-A-R-N vocabulary program to compile knowledge from URLs, files, or Brain artifacts into persistent knowledge packs. Use when ingesting new reference material or building knowledge packs. D

Reference for ~20 session, Guardian, and brain lifecycle hooks managing state, learning, and homeostasis. Use when looking up which hooks fire at SessionStart, Stop, or Compact events. Do NOT use for

Scan 8 system channels for threads that were started but never finished.

Enforce the Maat error equilibrium pattern for Rust APIs — target ~90% Ok / 10% Err paths (K_eq = 8.8). Use when reviewing error handling balance, auditing Result ratios, or designing new APIs with pr

Reference for MCP server development — transport config (stdio/SSE), registration patterns, tool schema design, and debugging. Use when building MCP servers, troubleshooting connections, or designing

Factory patterns for producing MCP servers, configs, proxy scripts, and tool schemas at scale. Use when generating new station tools, improving the dispatch pipeline, writing contract tests, or auditi

Audit nexcore workspace crates for MCP tool wiring coverage. Detects unwired domain crates, crate-tool mismatches, and consolidated server gaps. Use when checking which crates lack MCP exposure. Do NO

Run the M-E-N-D vocabulary program for closed-loop diagnostic remediation. Measure voids, Engineer fixes, Normalize across frameworks, Demonstrate reality score improvement. Use when diagnosing and re

Operate the microgram ecosystem — 22 commands for atomic self-testing programs. Generate, evolve, compose, stress, benchmark, patrol, and manage micrograms.

Analyze systems through 2+ complementary frameworks (Chemistry + Biology lenses) to produce a Unified Health Score and find cross-lens failure modes. Use when a single-perspective audit misses issues

Reference for NexCore REST API (port 3030) — endpoints, cross-domain signal detection, PVDSL, and Guardian APIs. Use when building or querying the nexcore-api server. Do NOT use for MCP tool developme

Validate the integration map between nexcore Rust workspace and Claude Code — MCP servers, hooks, skills, and transport boundaries. Use when diagnosing integration health or adding new wiring points.

Scan system state across 8 channels, score candidates on 5 dimensions, and surface ONE atomic action for permanent progress. Use when unsure what to work on or needing momentum direction. Do NOT use f

Next.js 16 App Router reference for Studio (port 9002) with 6 decision trees and 3 generator sub-skills. Use when building Studio pages, choosing component types, or configuring data fetching. Do NOT

Quick reference for querying openFDA REST APIs via mcp__nexcore__http_request.

Base Pair Rule — triage dangling bonds (open artifacts, unclosed loops). A=T, G≡C — no dangling bases in DNA. No dangling artifacts in sessions. Scans for TODO/BLOCKED/TBD artifacts, stale carry-forwa

Orchestrate parallel batch fixes across multiple failing deliverables using 6-phase protocol (AUDIT through CLOSE) with agent partitioning. Use when fixing batch failures across tests, docs, config, o

Enforce patient safety priorities (P0-P5) with ICH E2A seriousness escalation and Guardian triage matrix. Use when evaluating safety signals, triaging adverse events, or configuring escalation rules.

Apply performance mental models (Flow State, Decision Fatigue, Hick's Law, and 5 more) to optimize task execution and allocation. Use when diagnosing productivity bottlenecks or designing for flow. Do

Configure Claude Code output styles and persona profiles. Use when creating custom personas, defining tone profiles, or composing layered styles. Do NOT use for communication skills content — use comm

Run the P-I-X-E-L vocabulary program for visual design quality metrics. Use when measuring UI proportions, checking component hierarchy, or validating design system adherence. Do NOT use for accessibi

Strengthen points through presentation techniques — power periods, volume, connection, and presence (Ch6). Use when improving delivery strength or combating uptalk and weak endings. Do NOT use for poi

Define what a point IS and validate it with the three-step test (Ch1-Ch2). Use when formulating a clear contention or distinguishing topics from points. Do NOT use for delivery techniques — use point-

Convey and persuade effectively — make your point land, then sell it (Ch3-Ch4). Use when crafting value propositions, proposals, or persuasive arguments. Do NOT use for staying on point — use point-di

Stay on point, complete your delivery, and defeat the five enemies of communication (Ch5, Ch7, Ch8). Use when combating rambling, filler words, or weak conclusions. Do NOT use for point formulation —

Universal point-driven prompt that discovers context, identifies voids, respects experience boundaries, compounds KSBs, and uses config files as capability accelerators. Use when bootstrapping any ful

Apply point principles to specific media — speeches, PowerPoint, and email (Ch9). Use when preparing a talk, designing slides, or writing persuasive emails. Do NOT use for general point theory — use p

Compose sequential hook chains from HookPolymer patterns with stoichiometry and inter-molecule bonding. Use when building multi-hook execution pipelines, cyclic feedback loops, or defining hook sequen

Retrieve proteomics data from PRIDE/ProteomeXchange for protein extraction and quantification. Use when mining mass spectrometry datasets or validating protein-level hypotheses. Do NOT use for gene ex

Apply T1 Lex Primitiva operational SOP to decompose problems and compose solutions using the 16 axiomatic primitives. Use when performing primitive decomposition, composing solutions from first princi

Generate development procedures from topology routing. Takes a target node, decomposes into atomic transforms, composes a pipeline with verification gates, and outputs an executable procedure. Modeled

Run the P-R-O-G-R-E-S-S vocabulary program for measuring and producing real progress via Lex Primitiva. Use when verifying that work constitutes genuine progress (not churn, vapor, or drift), auditing

Official Claude Code hook protocol reference — exit codes, JSON I/O, decision patterns, hook types, and async behavior. Use when looking up protocol rules, exit code meanings, or hook I/O format. Do N

Run the P-U-L-S-E vocabulary program for system health monitoring. Use when checking service status, resource utilization, or threat signals across the AlgoVigilance stack. Do NOT use for code quality —

Detect perspective divergence in PV safety terms where author and reader hold different boundaries for the same word. Scans terms or narratives for perspective collisions that can suppress safety sign

Pharmacovigilance reference with signal detection formulas (PRR/ROR/IC/EBGM), Naranjo/WHO-UMC causality algorithms, and FAERS query patterns. Use when computing PV signals, assessing causality, or que

Run the 6-step quality pipeline with diagnostic scan, auto-repair, dependency audit, and GREEN/YELLOW/RED integration gate. Use when running pre-release quality checks or computing quality scores. Do

Railway Reference Model (RRM) — 6-layer safety-critical system architecture derived from CENELEC EN 50126/50128/50129 and ERTMS/ETCS. Classify components into layers, assign SIL, compute capacity, and

Browse curated tool and skill learning courses organized by role and domain. Use when onboarding to a domain, seeking progressive tool sequences, or asking "how do I learn X". Do NOT use for skill rec

Detect open loops, triage by impact, prescribe closures, execute with confirmation, verify they held, and self-check whether prior closures persisted.

Transform regulatory guidelines (ICH/FDA/EMA/CIOMS) into microgram decision trees through an 8-phase pipeline. Extracts IF-THEN rules, triages for codifiability, generates YAML micrograms, wires into

Extract and audit regulatory primitives from FDA/ICH/CIOMS/EMA guidelines with O(1) ICH glossary lookup (894+ terms). Use when analyzing regulatory text, extracting terms, or auditing guideline compli

Verify pipeline fidelity using formal relay theory — A1-A5 axiom compliance, fidelity composition (F_total, eta, alpha), and DF/CF/AF classification. Use when verifying pipeline correctness or detecti

Ingest research papers through the 7-phase Orient-to-Validate pipeline with fragment extraction, relevance triage (7+ gate), and crate BUILD/DO NOT BUILD assessment. Use when converting papers into ca

Orchestrate agent-driven research across data sources using three station modes (Field/Orbital/Polar), a 6-phase pipeline, and crew rotation patterns derived from physical research station design.

Apply resilience mental models (forgetting curve, antifragility, resource scarcity, recovery patterns) to system design. Use when addressing learning decay, designing for failure recovery, or analyzin

Detect rho (self-reference) failures where an artifact violates the constraint it enforces. Routes through two layers: Gate (automated, for artifacts with testable gates) and Probe (model-in-loop, for

ρ Gate — check if this session has encoded its own reader. DNA encodes the polymerase that reads DNA. Sessions must produce output that makes the next session smarter: verdict, memory, or carry-forwar

DNA Rhythm Protocol — session execution as polyrhythm. 5 rules mapped from DNA molecular architecture (source: session 77dacfd9, 2026-03-11). Routes to: arc (triplet tracking), exhale (stop codon), rh

Move from observed symptoms to underlying organizational issues using 5 Whys, Ishikawa categorization, and Theory of Constraints. Use when strategy gates fail, the same problem recurs, or a fix doesn'

Expert guidance on Rust architecture decisions and idiomatic patterns across seven anatomical levels. Use when making structural Rust decisions, choosing between patterns, or reviewing architecture. D

Capture and query Rust development learnings routed to brain system and rust-compounding.md. Use when recording a Rust insight, querying past learnings, or reviewing compounding progress. Do NOT use f

Rust development reference with 7 decision trees (type choice, Clone/Copy, ownership, smart pointers, interior mutability, error handling, async) and 18 sub-skill router. Use when writing Rust code, c

Route Rust compiler errors to the correct diagnostic sub-skill by error category (535+ error codes, Edition 2024). Use when encountering rustc errors like E0xxx, "cannot move", "mismatched types", or

Diagnose advanced Rust errors: const/static, structs/enums, FFI/unsafe, visibility, generics, Send/Sync. Use when encountering errors in these categories. Do NOT use for ownership errors — use rust-er

Diagnose Rust control flow errors: closures, async, modules, imports, and pattern matching. Use when encountering closure capture, async lifetime, or module resolution errors. Do NOT use for type erro

Diagnose Rust ownership, borrowing, and lifetime errors with fix patterns. Use when encountering "cannot move", "cannot borrow", or lifetime mismatch errors. Do NOT use for type errors — use rust-erro

Quick lookup tables for Rust error codes, complete error index, NexCore-specific notes, and Edition 2024 changes. Use when looking up a specific error code or checking Edition 2024 migration. Do NOT u

Diagnose Rust type system and trait implementation errors with fix patterns. Use when encountering "mismatched types", "trait bound not satisfied", or "expected type" errors. Do NOT use for ownership

Enforce Rust safety via 10 bash guardrail hooks (unwrap guardian, panic detector, transmute blocker, clippy enforcer, etc). Use when checking which Rust hooks are active or understanding enforcement r

Reference for Rust 2024 Edition changes: RPIT capture rules, unsafe extern, static_mut_refs, resolver v3, gen keyword, and migration guide. Use when migrating to Edition 2024 or understanding new lang

Reference for Cargo manifest, dependencies, workspaces, profiles, build scripts, features, registries, and commands. Use when configuring Cargo.toml, managing workspace deps, or understanding build pr

Reference for building CLI apps in Rust with clap, error handling, output formatting, testing, signals, exit codes, and distribution. Use when building command-line tools in Rust. Do NOT use for gener

Reference for embedded Rust development: #![no_std], PAC/HAL/Board layers, interrupts, concurrency, and memory-mapped I/O. Use when writing firmware or working with microcontrollers in Rust. Do NOT us

Comprehensive Rust compiler error code reference (E0001-E0805) with diagnosis patterns, fix templates, and NexCore-specific guidance. Use when looking up a specific E0xxx error code. Do NOT use for er

Reference for rustc compiler flags, codegen options, lint system, target tiers, PGO, and LTO configuration. Use when configuring compiler options, tuning codegen, or understanding target tiers. Do NOT

Reference for Rust documentation: doc comments, #[doc] attributes, doctests, intra-doc links, and doc sections. Use when writing rustdoc comments, configuring doc tests, or using intra-doc links. Do N

Reference for Rust standard library: key modules, types, traits, and module hierarchy for std/core/alloc. Use when looking up std types, trait implementations, or module organization. Do NOT use for e

Scaffold Rust modules and crates with governance using the module-vs-crate decision tree and pre-creation checklist. Use when deciding whether to create a module or crate, or scaffolding module struct

Mine FAERS safety data to identify exploitable safety gaps in incumbent drugs, model molecular redesigns, and war-game competitive entry using Nash equilibrium. Use when exploring reverse PV drug desi

Run the S-C-O-P-E vocabulary program for project analysis and profiling. Use when scanning project structure, counting metrics, outlining architecture, or profiling quality. Do NOT use for code qualit

Capture and analyze the user's screen. Three modes — full (max detail), efficient (balanced, default), thumb (change detection). Use when user asks to see their screen, verify UI, or when visual conte

Apply the 5-gate diagnostic conservation law to the model's own prior output. Finds bold claims, unsupported assertions, premature convergence signals, and apophenic pattern-matching in recent respons

Reference for Claude Code settings and configuration — settings.json, CLAUDE.md memory files, permissions, environment variables, and config precedence. Use when configuring Claude Code behavior or un

Cross-language guardrail hooks applying to both Rust and TypeScript files (secret scanner, SQL injection, command injection, bloat detector, doc density, brand enforcer). Use when checking which share

Recommend relevant skills for the current context with confidence scoring and session memory. Use when unsure which skill to use, starting a new task, or wanting skill suggestions. Do NOT use for skil

Assess skill ecosystem health, compliance levels, and sync gaps across all skills. Use when auditing the skill fleet, checking compliance distribution, or identifying missing agents. Do NOT use for in

Guide for creating and updating effective Claude Code skills with proper frontmatter, structure, and testing. Use when building a new skill from scratch or optimizing an existing skill's performance.

Manage skill lifecycle across 5 modes (CREATE/VALIDATE/AUDIT/INTERACTIVE/HYBRID) with official frontmatter spec and compliance levels (Bronze to Diamond). Use when creating, validating, or auditing sk

Install skills from curated list or GitHub repos into the skills directory. Use when adding new skills from external sources or listing available installable skills. Do NOT use for skill creation — us

Test skill effectiveness by measuring input/output quality across task variety. Use when scoring skill quality, running effectiveness tests, or identifying coverage gaps. Do NOT use for skill creation

Set, refine, score, and troubleshoot SMART goals with milestones and KPIs. Use when setting goals, diagnosing why goals failed, or refining objectives to be specific and measurable. Do NOT use for pro

Apply social psychology mental models (bystander effect, social proof, groupthink, and more) to Claude Code tool behavior and team dynamics. Use when analyzing group behavior patterns or designing for

Teach through methodical decomposition with intellectual humility and genuine curiosity using the Socratic method. Use when explaining complex topics step-by-step or guiding discovery through question

Assess AlgoVigilance SOPs against HIPAA, GDPR, and 21 CFR Part 11 compliance frameworks with gap analysis and remediation guidance. Use when checking regulatory compliance or preparing for audits. Do NO

Generate new AlgoVigilance SOPs from master template with all 18 governance sections populated. Use when creating a new SOP from scratch or scaffolding governance documentation. Do NOT use for SOP valid

Inject SOP governance sections into any skill's SKILL.md by analyzing purpose and phases. Use when embedding procedure governance into skills or standardizing skill execution. Do NOT use for standalon

AlgoVigilance SOP governance framework for generating, validating, and managing Standard Operating Procedures compliant with HIPAA, GDPR, and 21 CFR Part 11. Use when managing the SOP lifecycle or estab

Validate AlgoVigilance SOPs against 18-section governance requirements for completeness, formatting, and structural integrity. Use when auditing an existing SOP or verifying governance compliance. Do NO

Eliminate external dependencies by building minimal sovereign replacements, then migrate all consumers atomically. Use when a dependency poses license, security, or control risk. Do NOT use for genera

Identify which of the 6 core Stark cognitive patterns to apply (Ch1). Use when starting a creative session or unsure which thinking blade to use. Do NOT use for executing a technique — use the specifi

Apply advanced Stark techniques (Complexity Assessment, Parallel Hypothesis, Stakeholder Map, Reversibility Analysis, Systems Stress Test, Evolution Architecture) for complex multi-variable problems.

Configure your mental workshop before executing (Ch2) — daily practices, block breakers, and Cave Exercise. Use when starting a creative session or feeling blocked. Do NOT use for problem-solving — us

Walk through a guided ITERATE cycle on a real problem (Ch3). Use when wanting a structured problem-solving walkthrough. Do NOT use for theory — use stark-ch1-mindset instead. Triggers on '/stark-chall

Dig through 5 assumption layers using the Archaeology method (Ch4). Use when deep assumption analysis is needed or surface solutions keep failing. Do NOT use for rapid prototyping — use stark-ch7-rapi

Find connections across disciplines for breakthrough innovation (Ch5) with Random Collision, Expert Perspective, and Metaphor Bridging. Use when cross-domain insight is needed. Do NOT use for first-pr

Rapid prototype with the 3-2-1 Rule, Learning Loop, and Kill Criteria (Ch7). Use when building and testing ideas quickly. Do NOT use for deep analysis — use stark-ch4 or stark-ch11. Triggers on '/rapi

Turn failure into a research method (Ch8) with Failure Design Canvas, Autopsy, Five Whys, and Failure Budget. Use when learning from mistakes or designing safe experiments. Do NOT use for success plan

Map systems with Component Inventory, Connection Mapper, Feedback Loops, Leverage Points, and Transaction Trace (Ch9). Use when understanding how everything connects. Do NOT use for simple architectur

Apply the Stark Method — ITERATE (Identify, Think, Experiment, Refine, Architect, Test, Evolve) with cross-disciplinary innovation and first principles. Use for breakthrough problems that resist conve

AlgoVigilance Station tool demonstration with real sample inputs across all 16 config domains. Run individual tools or the Quick Demo workflow for a mini PV signal detection pipeline. Triggers on 'stati

Run end-to-end PV signal detection pipeline using AlgoVigilance Station MCP tools. Resolves drug identity, queries FAERS, computes disproportionality (PRR/ROR), checks DailyMed labeling, searches PubMed

Compute STEM-grounded values that anchor mental models to measurable empirical data. Use when quantifying abstract concepts, validating thresholds, or grounding qualitative assessments. Do NOT use for

Plan strategy using the Playing to Win framework with capability analysis and management system design. Use when making strategic decisions, building roadmaps, or analyzing competitive position. Do NO

Separate primitives needed for existence from primitives needed for success. After subtractive convergence, classify survivors as structural-only (needed to exist) or structural+strategic (needed to w

Run the 4-gate AlgoVigilance site development strategy review. Checks Academy→Glass, Microgram→Station wiring, Worked Example, and External Traffic gates in dependency order. Derived from 20Q Scramble (

Pass a strict 23-lint deny clippy configuration in a single pass. Use when preparing Rust code for the strict clippy gate or fixing as_conversions, collapsible_if, safety trio, and determinism lints.

Generate Next.js API routes connecting Studio to NexCore with proper auth and proxy patterns. Use when creating new API endpoints for Studio. Do NOT use for page generation — use studio-page-gen inste

Generate Studio components following shadcn/ui patterns with AlgoVigilance design tokens. Use when creating new UI components for Studio. Do NOT use for page generation — use studio-page-gen instead. Tr

Generate new Studio pages following existing patterns at ~/Projects/Active/nucleus/. Use when creating new pages for the Studio web app. Do NOT use for components — use studio-component-gen instead. T

Reference for subagent configuration — task delegation, context isolation, model selection, and skill preloading. Use when designing agent architectures or configuring subagent behavior. Do NOT use fo

Find the essential primitives of any concept by iterative elimination. Apply conservation gates, remove what fails, repeat until removal breaks the concept. What survives IS the definition.

Orchestrate system operations via the 7-phase CONTROL DAG with Guardian PAMP/DAMP threat model and MCP tool routing. Use when managing system health, coordinating teams, or responding to threats. Do N

Measure current AlgoVigilance system state across all surfaces. Loads the system-status-prompt from brain and executes the full 13-surface + 6-wire + primitive reading + conservation check cycle. Use at

Apply systems dynamics mental models (Cognitive Dissonance, Gaussian Distribution, Kubler-Ross, and 5 more) to complex system behavior. Use when analyzing group dynamics, boundary conditions, or emerg

Activate metacognitive systems thinking with feedback-loop awareness, assumption surfacing, and primitive-based reasoning. Use when facing complex problems requiring structural analysis. Do NOT use fo

Quick reference for Tailwind CSS and the Golden Ratio design system in Studio. Use when styling components, choosing colors, spacing, typography, or writing CVA variants. Do NOT use for accessibility

Orchestrate pre-configured agent teams from 11 templates (CRAFT/GUARD/FORGE/STUDIO/PULSE/LEARN/TRACE/SCOPE/MEDIC/FOUNDRY/DIRECTIVE). Use when spawning multi-agent teams for parallel work. Do NOT use f

Governance layer for agent teams with 7 specialist capabilities and composable patterns (Lean/Hardened/Full Spectrum/Hybrid). Use when designing team structures or defining quality gates. Do NOT use f

Reference for Rust testing methodology — unit, integration, property, fuzz, and coverage analysis. Use when organizing tests, writing assertions, or planning test strategy. Do NOT use for test executi

Enforce zero-tolerance Rust test protocol with 4 non-negotiable rules. Use when running tests, investigating failures, or ensuring test discipline. Do NOT use for test design — use test-dev instead. T

Look up PreToolUse tool_input schemas for every built-in Claude Code tool. Use when writing hooks that inspect tool inputs or building PreToolUse matchers. Do NOT use for hook protocol — use protocol

Run the T-R-A-C-E vocabulary program for debug tracing and error analysis. Use when debugging errors, tracing failure paths, or analyzing error patterns across sessions. Do NOT use for system health —

Compute cross-domain transfer confidence using the 3D formula (structural x 0.4 + functional x 0.4 + contextual x 0.2). Use when scoring T1/T2/T3 primitive transfers or assessing mapping reliability.

Execute cross-domain transfers by decomposing methods to T1 primitives, computing resonance, and generating adapted versions. Use when applying a technique from one domain to another. Do NOT use for c

Run the T-R-I-A-L vocabulary program for hypothesis-driven experimentation. Use when designing experiments, running A/B tests, or validating models with FDA clinical trial methodology. Do NOT use for

Enforce TypeScript safety via 7 bash guardrail hooks (any blocker, assertion blocker, console blocker, tsc verifier, etc). Use when checking which TS hooks are active or understanding enforcement. Do

Play 20 Questions (decomposition) or Anti-20 Questions (composition) to systematically break down or build up concepts using T1 primitive navigation. Use when user says '/20q', '/anti-20q', '/twenty-q

Run adaptive multi-target validation with 8 gates for code (Rust/TypeScript), Jupyter notebooks (cohesive front-to-back read), and Claude Code artifacts (skills, agents, hooks, CLAUDE.md). Use when va

Orchestrate validated SMART-DAG pipelines integrating CTVP, CEP, and SECI for goal validation and closed-loop learning. Use when running complex multi-phase goal execution with validation. Do NOT use

Activate the Vigil CAIO executive system for AI organizational management. Use when running business simulation, managing agent fleet, or generating executive reports. Do NOT use for PV signals — use

Reference for NexCore FRIDAY (Vigil) orchestrator — voice, webhooks, scheduler, event bus, and decision engine. Use when building orchestration, event-driven systems, or voice integration. Do NOT use

NexCore Vigilance reference with 25+ domain modules, 5 ToV axiom definitions, 8-type harm taxonomy (A-H), and Guardian-AV risk scoring. Use when working with PV domain modules, ToV axioms, or harm cla

Route to the correct vocabulary program from the 15 available (CRAFT, GUARD, CLEAN, PROGRESS, etc). Use when unsure which program to run or asking 'which program'. Do NOT use for individual programs —

Curate and maintain the shared vocabulary system — review proposals, add shorthands, prune unused, audit health. Use when managing constraint compression vocabulary. Do NOT use for vocabulary program

Scan the skill ecosystem for capability gaps, map to operational layers, and produce a prioritized Void Map. Use when identifying what to build next or auditing coverage. Do NOT use for skill quality

Start a Voila notebook + headed Playwright browser testing session. Serves a notebook via HTTPS, launches a visible browser, and executes the user's objective.

Apply the fearless Rust development doctrine with battle metaphors for memory safety ownership patterns. Use when approaching complex ownership challenges or building confidence with the borrow checke

Continuous delta-aware screen monitoring. Captures frames only when the screen changes — cursor blink and compositor noise filtered out. Token-efficient persistent visual attention.

Process incoming webhook events — validate signatures, transform payloads, persist to event store, and notify subscribers. Use when building webhook handlers or event-driven pipelines. Do NOT use for

Query WebMCP Hub for community-contributed browser automation configs, generate automation code, and translate site interaction patterns into Rust or Playwright implementations. Use when user says '/w

Scan the 3-surface architecture (Anatomy/Physiology/Nervous System) and produce a connectivity map showing which surfaces are wired, which are void, and the overall wiring coverage ratio. Feeds flywhe