Claude Skills by curiositech

Creates realistic cloud effects for web using SVG filters (feTurbulence, feDisplacementMap), CSS animations, and layering techniques. Use for atmospheric backgrounds, weather effects, skyboxes, parallax scenes, and decorative cloud elements. Activate on "cloud effect", "SVG clouds", "realistic clouds", "atmospheric background", "sky animation", "feTurbulence", "weather effects", "parallax clouds". NOT for 3D rendering (use WebGL/Three.js skills), photo manipulation (use image editing tools), ...

Creates unique web designs with brand identity, color palettes, typography, and modern UI/UX patterns. Use for brand identity development, visual design systems, layout composition, and responsive web design. Activate on "web design", "brand identity", "color palette", "UI design", "visual design", "layout". NOT for typography details (use typography-expert), color theory deep-dives (use color-theory-expert), design system tokens (use design-system-creator), or code implementation without des...

Build and operate a small-business storefront end-to-end — choose the platform route (Stripe/Snipcart/Shopify/merchant-of-record), integrate checkout into an existing React/Tailwind/Radix site so it matches the design, wire payouts to the owner's bank, design SKUs and a catalog-as-code system with sync scripts, cross-list to Google/Meta/Etsy/eBay via feeds and APIs, render static HTML previews in the site's own style, and hand off operator README/FAQ/update-playbook docs. Use for "add a shop/...

Creates realistic ocean and water wave effects for web using SVG filters (feTurbulence, feDisplacementMap), CSS animations, and layering techniques. Use for ocean backgrounds, underwater distortion, beach scenes, ripple effects, liquid glass, and water-themed UI. Activate on "ocean wave", "water effect", "SVG water", "ripple animation", "underwater distortion", "liquid glass", "wave animation", "feTurbulence water", "beach waves", "sea foam". NOT for 3D ocean simulation (use WebGL/Three.js), ...

Create stylized SVG and CSS weather effects for web interfaces, with layered motion, visual restraint, and accessible performance choices.

Implementing the relying-party (server) side of WebAuthn / passkeys to replace passwords — the registration ceremony, the authentication ceremony, the verification steps the server MUST perform on each (challenge match, origin match, RP ID hash, signature, counter monotonicity), conditional UI for autofill, and the synced-vs-device-bound distinction. Grounded in W3C WebAuthn-3, FIDO Alliance, passkeys.dev, and SimpleWebAuthn. NOT for: CTAP authenticator internals, OAuth/OIDC layering, TOTP/SM...

Use when designing webhook endpoints (Stripe/GitHub/Slack/internal), implementing HMAC signature verification, ensuring idempotency on retries, handling out-of-order events, building dead-letter queues, or replaying historical webhooks. Triggers: signature header verification (Stripe-Signature, X-Hub-Signature-256), timestamp window for replay protection, idempotency keys, exactly-once processing semantics, retry storms, webhook handler latency budgets, async vs sync processing, dead-letter c...

Transform thousands of wedding photos and hours of footage into an immersive 3D Gaussian Splatting experience with theatre mode replay, face-clustered guest roster, and AI-curated best photos per person. Expert in 3DGS pipelines, face clustering, aesthetic scoring, and adaptive design matching the couple's wedding theme (disco, rustic, modern, LGBTQ+ celebrations). Activate on "wedding photos", "wedding video", "3D wedding", "Gaussian Splatting wedding", "wedding memory", "wedding immortalize...

Strategic framework for eliciting and orchestrating reasoning in LLM-based agents through structured decomposition

Defensive security and mechanization fleet for the Port Daddy whitepapers (Bonded Commons, Anchor Protocol). Use when responding to red-team findings in a versioned round, when closing a cited-but-unmodeled proof, or when proposing the next paper version bump. NOT for ad-hoc code review — see code-reviewer skill.

Diagnoses, selects, and audits the semantic form of technical-paper figures before TikZ implementation. Use for any figure in the seven Port Daddy whitepapers, especially when the current result is generic boxes and arrows, prose in shapes, an ornamental illustration, visually homogeneous nodes, a misleading chart type, or a layout that does not reveal the paper's claim. Pair with tikz-figure-engineering for source, rendering, page-fit, and contact sheets. NOT for cover or chapter art, statis...

Build WinDAGs — the orchestration platform where AI agents accumulate genuine expertise through DAGs of skillful agents. Covers DAG design, execution engines, meta-DAG architecture, skill selection, dynamic mutation, visualization, and deployment. Activate on "windags", "agent DAG", "DAG of agents", "workflow orchestration", "agent pipeline", "dynamic DAG", "meta-DAG", "build windags", "implement windags". NOT for understanding WHY decisions were made (use windags-avatar), creating individual...

The living embodiment of WinDAGs V3 — the AI orchestration platform where agents accumulate genuine expertise. Knows the full constitution, convention process, tradition positions, behavioral contracts, and derivative documents. Advises on architecture with ADR provenance, explains decisions with tradition attribution, grows source material as we build. Activate on "windags", "WinDAGs", "constitution", "convention", "why did we decide", "what does WinDAGs say", "how should we build", "archite...

Post-execution skill crystallization and learning engine updates for WinDAGs. Runs after successful execution to update Thompson sampling parameters, track method quality, detect monster-barring, log near-miss events, and signal Kuhnian crises. Activate on "curator", "learning update", "skill crystallization", "Thompson sampling", "monster-barring", "near-miss", "Kuhnian crisis", "post-execution learning". NOT for pre-execution risk scanning (use windags-premortem), retrospective analysis (us...

Deep understanding of WinDAGs customer personas — who needs DAG-based agent orchestration, their pain points, buying triggers, and workflow patterns. Activate on 'who is the WinDAGs customer', 'target audience', 'customer persona', 'ICP', 'ideal customer profile', 'who buys agent orchestration', 'WinDAGs positioning', 'competitive differentiation'. NOT for marketing execution (use next-move-marketing), general competitive mapping (use competitive-cartographer), or pricing strategy (use indie-...

Maintain consistency across all winDAGs documentation surfaces — corpus documents, marketing copy, README files, site content, skill files, API docs, and derivative documents. Enforces the canonical terminology table, voice guidelines, architectural accuracy, and cross-reference integrity. Activate on "windags docs", "documentation consistency", "windags copy", "terminology check", "docs audit", "windags readme", "site copy", "docs sync", "content review", "windags librarian". NOT for writing...

Runs CDM (Critical Decision Method) interviews to elicit the perceptual layer missing from science-only SKILL.md files. Activate when: Looking Back Q3 fires with scope≠narrow, a pattern_crystallization event is detected, a skill underperforms (expectancy_violation), or a novel task has no catalog match. Produces structured SKILL.md v2 metadata: recognition-cues, expectancies, decision-cues, adaptive-workarounds. NOT for: general Q&A, skill selection, DAG execution.

Stage 3a of the WinDAGs meta-DAG. Given a decomposed subtask graph and a pre-narrowed candidate list per subtask, picks a primary and runner-up skill, plus model tier and cost estimates. Activate on "skill selection", "skill matching", "pick skills for subtasks", "skill cascade", "primary and runner-up", "model tier selection". NOT for pre-narrowing candidates (the SkillSearchService cascade does that), generating subtasks (use windags-decomposer), or building a runnable PredictedDAG (use win...

Stage 4 (final) of the WinDAGs meta-DAG. Assembles the final PredictedDAG from Sensemaker, Decomposer, Skill Selector, and PreMortem outputs. Picks planning topology, names runtime divergence honestly, validates against the schema, and emits a runnable plan. Activate on "synthesize predicted DAG", "assemble final plan", "topology selection", "runtime honesty", "predict assembly". NOT for generating subtasks (windags-decomposer), classifying problems (windags-sensemaker), picking skills (winda...

Modern web applications with authentic Windows 95 aesthetic. Gradient title bars, Start menu paradigm, taskbar patterns, 3D beveled chrome. Extrapolates Win95 to AI chatbots, mobile UIs, responsive layouts. Activate on 'windows 95', 'win95', 'start menu', 'taskbar', 'retro desktop', '95 aesthetic', 'clippy'. NOT for Windows 3.1 (use windows-3-1-web-designer), vaporwave/synthwave, macOS, flat design.

Longitudinal memory tracking, philosophy teaching, and personal accountability with compassion. Expert in pattern recognition, Stoicism/Buddhism, and growth guidance. Activate on 'accountability', 'philosophy', 'Stoicism', 'Buddhism', 'personal growth', 'commitment tracking', 'wisdom teaching'. NOT for therapy or mental health treatment (refer to professionals), crisis intervention, or replacing professional coaching credentials.

Comprehensive introduction to multi-agent systems covering coordination, communication, and distributed AI theory

Capture one WorkIntent, record six local topology signals with reasons, select one local work shape, and audit supplied route-trace evidence before admitting the canonical WorkIntent to WorkPlan pipeline. Use when a request may need a node, scout, chain, DAG workgroup, tournament, watcher, or human gate. NOT for universal topology classification, proving a route never writes state from a boolean alone, or preserving a legacy launch route by default.

Design bounded AutoGen 0.2 conversations with explicit computation, control, execution, and human-input boundaries. NOT for assuming code execution, human approval, or termination is safe by default.

Comprehensive survey of LLM-based task planning methods including decomposition, search, and execution strategies

Use ReAct's interleaved action and observation pattern for evidence-carrying tool work. NOT for treating model reasoning traces as evidence without checking the tool observations.

Use explicit domain search states, evaluation, and backtracking for bounded LLM-assisted search; NOT for ordinary one-shot prompting without branching or state checks.

Use when changing schema on a live production database without dropping traffic, doing expand/contract migrations, backfilling new columns on million-plus-row tables, dual-writing across schemas, renaming a hot column, splitting a table, or adding a foreign key to a big table. Triggers: ALTER TABLE locks the world, "must be removed in a separate deploy", expand-contract-cleanup phases, NOT NULL on populated column, foreign-key add on big table, dropping a column still read by old replicas, lo...

Bounded neuro-symbolic oversight research using Dafny/Z3; not a guarantee of agent correctness or deployed enforcement

Handles alpha catalog tasks for reconciler tests.

Handles beta catalog tasks for reconciler tests.

Handles gamma catalog tasks for reconciler tests.

Designs and audits deterministic, authority-filtered partitions of trust-typed context for already admitted bodies or abstract continuation slots. Use when causal context, obligations, disclosure boundaries, vector-space identity, capacity, and omission proofs must survive partitioning. NOT for spawning or admitting agents, choosing worker count, retrieving arbitrary knowledge, writing successor prompts, or granting tools, leases, identity, or effect authority.

Designs and audits closed, ordered, replay-safe message protocols between already admitted agent bodies. Use when a multi-agent exchange needs explicit epochs, audiences, per-sender sequencing, gathers, terminal fences, acknowledgements, and deterministic replay. NOT for selecting or spawning agents, granting identity or authority, choosing a runtime topology, general human facilitation, wire transport implementation, or deciding whether evidence is true.

Design AgentSpeak(L)-style BDI agents with context-guarded plans, selection functions, and intention stacks. Use for interruptible autonomy, agent policy, and multi-agent orchestration in dynamic environments. NOT for simple rule engines, static planners, or centralized workflows.

Designs message-driven actor interactions with explicit delivery, recovery, and capability assumptions. NOT a transport reliability or authorization guarantee.

Apply AlphaGo-style patterns: learned priors plus search, mixed evaluators, self-play, and asynchronous fast/slow loops. Use for intractable search and compute allocation. NOT for basic supervised learning, tractable optimization, or domains without search bottlenecks.

Model surveillance knowledge only with explicit camera geometry, observation, action, and epistemic assumptions.

Apply the Vasiliauskaite, Evans and Expert cycle-basis/orientation method to a valid DAG with explicit ordering metadata. Use for graph-structural descriptors; not SCC-only diagnosis, scheduling, or resilience proof.

Separates agent identity, addresses, lifecycle, and directory discovery using FIPA agent-management concepts. NOT a transport, reachability, or authorization guarantee.

Specifies role-based interaction protocols with explicit conversation state and local transport/failure overlays. NOT a conformance, reliability, or authorization guarantee.

Simulation of believable human behavior using LLM-powered generative agents with memory and social interaction

Pre-decomposition review of task definition, constraints, authority, and feasibility. Uses typed hard blockers and decision-relevant clarification; any scalar score or cutoff is an implementation-specific policy that requires its own calibration and error-cost rationale. Avoids presenting Polya's problem-analysis framework as a source for numeric weights or thresholds. The gate is named "pre-federation" because it fires before agents federate around a shared decomposition — once subtasks exis...

Designs and audits Drydock-style hostile-execution laboratories for untrusted, AI-authored, or pull-request-controlled code, including VM isolation, brokered I/O, provider spend custody, deterministic simulation, sealed provenance, external receipts, trusted-language/process selection, durable agent lifecycle accounting, crash-storm breakers, fail-cheap controls, and promotion gates. Use before a daemon, agent, test suite, build, or skill may touch real machines, networks, credentials, provid...

AST-based semantic conflict prediction for multi-agent coding environments. Uses tree-sitter to parse code into ASTs, extract symbol-level claims, build dependency graphs, and predict conflicts BEFORE they happen -- at work-assignment time, not merge time. The key insight: git's textual merge is necessary but not sufficient. Two changes can merge cleanly and break the program. This skill covers tree-sitter fundamentals, symbol-level claims, dependency graph construction, conflict prediction a...

Design bounded delegated workers by specifying outputs, inputs, tools, effects, stop rules, evidence, and handoffs. NOT a claim that prompts enforce permissions or that fixed skill counts, sections, or question quotas are optimal.

Uses Smith's 1980 Contract Net Protocol for local task announcement, bid, award, and result coordination. NOT FIPA Contract Net, reliable transport, or automatic authority to execute.

Design one accountable coordinator for a bounded concern, with durable handover and independently controlled effects.

Design durable local SQLite state for multi-agent developer tooling that survives package upgrades, concurrent writers, and crashes: canonical env-pinned paths, WAL/busy_timeout discipline, idempotent verified migrations, and safe writer topology. Use when a daemon, CLI, or agent fleet needs a persistent local DB, when reviewing a schema/path change before it ships, or when diagnosing an already-fragmented multi-.db mess. NOT for relational schema/query design (use daemon-development for the ...

Analyze human-automation work as situated activity, then test recovery and observability on representative scenarios.