AI & ML

thewatcher01 Bundle

Set up and maintain SOTA agentic architectures in VS Code workspaces with multiple projects. Use when: setting up agents, agentic architecture, multi-project workspace, agent hierarchy, context engineering, prompt engineering, skills ecosystem, AGENTS.md, copilot instructions, agent handoffs, orchestrator pattern, progressive disclosure, workspace agent structure. Teaches agent taxonomy design, skill ecosystems, instruction layering, handoff graphs following anthropics/skills, OpenAI Agents SDK, and GitHub Copilot custom agent standards (VS Code 1.106+).

thewatcher01 Skill

Systematic AI-powered code review. USE when: reviewing PRs, auditing code quality, checking for security vulnerabilities, validating architecture, analyzing Rust/Python/JS/TS code, reviewing config files, or when the user says "review this code", "check this PR", "find bugs", "is this code safe". Covers: security (OWASP), performance, maintainability, correctness. Optimized for fast review with Haiku 4.5.

thewatcher01 Skill

Configure and use free LLM providers with intelligent rotation and fallback. USE when: setting up ZeroClaw with free tiers, hitting rate limits, switching providers dynamically, comparing free model quality, configuring Ollama local fallback, or when the user says "use a free model", "I hit the rate limit", "switch to a cheaper model", "configure Ollama". Covers OpenRouter (26+ free), Groq (free tier), Ollama (local), Cerebras.

thewatcher01 Bundle

Expert guide for creating state-of-the-art chainskills workflows following MCP and agent skills standards. Use when designing multi-step agent workflows, converting processes to .workflow.md format, or building production-grade automation pipelines.

thewatcher01 Skill

Optimize the AI context window for any task. USE when: agent has poor memory or forgets context, responses are imprecise, needing to compress long conversations, building RAG pipelines, designing multi-turn agent state, managing tool results in context, or when the user says "the AI doesn't remember" or "context too long". Implements Karpathy/Willison context engineering: right info, right time, right format, right size.

thewatcher01 Skill

Security audit for agentic systems, MCP servers, skills, hooks, workflows, and external tools. USE when: adding any external skill/MCP/plugin/tool, reviewing agent configs, checking hook code, auditing workflows, detecting prompt injection, validating secrets management, reviewing agent permissions. CRITICAL: "never reproduce the openclaw scandal". Covers OWASP Top 10 for AI agents + supply chain security. Trigger: "audit this skill", "is this MCP safe", "check agent security", "review this tool".

fusengine Bundle

Use when auditing an existing prompt for clarity, structure, or completeness issues, fixing vague instructions, or producing a before/after optimization report.

kevinnft Skill

MCP client: connect servers, register tools (stdio/HTTP).

kevinnft Skill

Karpathy's LLM Wiki: build/query interlinked markdown KB.

fusengine Bundle

Use when writing or reviewing modern TS syntax — const type parameters, using/await using, standard decorators, satisfies, or LLM anti-patterns (legacy enum/namespace). Not for tsconfig (ts-config).

fusengine Bundle

Use when deciding where TanStack Start code runs, fixing secret leaks, DB-in-loader bugs, or hydration mismatches.

jacob-balslev Skill

Use when working with Etsy listings — Etsy's own title/tag/category/attribute/description/photo/video model and two-phase search (query matching → ranking on relevancy, listing engagement, recency, shipping price, customer-and-market experience, and personalization); the 13-tag and 20-character-per-tag limits and 140-character title cap; Etsy's 2026 natural-readable-title guidance; the description as an indexed/semantic and Google-snippet surface; free-shipping/shipping-price visibility thresholds; listing-video specs; the first-party Search Visibility diagnostic; the Etsy Open API listing surface; and editing Etsy listings managed by an upstream POD platform (Printify/Printful) including the publish/lock lifecycle, selective publishing, and publish-failure diagnostics. Do NOT use for generic Shopify storefront work, non-Etsy marketplaces, deep keyword-research methodology, the POD product/blueprint/variant model itself, or print-file/artwork generation.

x402agent Skill

Time-Weighted AMM crank automation for the Clawd agent. Use when scheduling large spot orders to execute at the time-weighted oracle price with near-zero slippage, operating the permissionless on-chain TWAMM crank program.

x402agent Skill

Phoenix on-chain market-maker automation for the Clawd agent. Use when running the Rust mm binary that places resting limit orders on the Phoenix CLOB to earn maker rebates and tighten the book spread.

x402agent Skill

MUST be loaded before producing any answer about Vulcan, Phoenix DEX, Phoenix perpetuals, Solana perps, the `vulcan` CLI/MCP, or agent-based trading workflows. Single entry-point: contains the runtime contract pointer, the non-negotiable safety rules, the focused-skill router, and the live-launch preflight gate. Answers given without loading this skill have produced wrong commands, wrong wallet selection, and unsafe live trades. Load on the FIRST turn the topic appears in a conversation and re-load on any session restart.

jacob-balslev Skill

Use when deciding whether to run a task in the Codex CLI agent harness (which drives a frontier GPT model), when scoping work to its native capabilities (resumable `codex exec resume` sessions, non-interactive `exec` stdout-piping, in-process dispatch, `/permissions` sandbox modes, MCP, on-demand subagents, the `/review` code-review agent), or when choosing Codex versus the Claude Code harness for a piece of work — and when avoiding its known failure modes (non-autonomous subagents, Full-Access network reach, cold-one-shot context loss). Do NOT use for routing a task to the GPT MODEL versus Claude (use `gpt-5-5`), for the Claude Code harness decision (use `claude-code`), or for designing a generic resumable agent loop (use `autonomous-loop-patterns`). Do NOT use for is GPT-5.5 or Opus the better model for this task? Do NOT use for what is Claude Code good at? Do NOT use for design a resumable supervised loop from scratch.

jacob-balslev Skill

Use when deciding whether to route a task to OpenAI's GPT-5.5 frontier model versus Claude Opus or Sonnet — picking the model lane for infrastructure scripts, CI pipelines, concrete implementation, analytical code review, security review, or CLI/terminal-heavy work, and weighing GPT-5.5's context window, pricing, and per-benchmark strengths against the Claude tiers for the same task. Covers the decision-useful capability and pricing facts, and the boundary against the Claude routing skills. Do NOT use for running the GPT model through a harness (use `codex`), for choosing among the Claude tiers themselves (use `claude-opus` / `claude-sonnet`), or for routing among local skills at request time (use `skill-router`). Do NOT use for resume my last Codex session and keep going. Do NOT use for is Opus or Sonnet the right Claude tier for this? Do NOT use for which of my skills handles webhook tasks?

jacob-balslev Skill

Use when deciding whether to run a task on the OpenCode agent runtime, how to invoke it non-interactively, how to pick a provider/model string, or how OpenCode differs from Claude Code and Codex. Covers terminal TUI, opencode run, ACP/IDE bridge, desktop beta, web UI, opencode serve, provider/model routing, OpenCode Zen and OpenCode Go lanes, config precedence, JSONL output, CLI command surface, build/plan agents, LSP code intelligence, permissions, Agent Skills, MCP, references, commands, plugins, local models, and scripting/automation. Do NOT use for choosing which free model fits a task (use opencode-free-models), writing the agent loop itself (use autonomous-loop-patterns), or GitHub Copilot premium-request economics (use github-copilot). Do NOT use for which free model should I use for this classification job? Do NOT use for how do I structure the autonomous agent loop itself? Do NOT use for how many Copilot premium requests will this burn?

jacob-balslev Skill

Use when working with Printify — the print-on-demand REST API, catalog model (blueprints, print providers, variants, print areas), product creation and publish lifecycle to connected channels, order routing, shipping cost queries, and HMAC SHA-256 webhook verification. Do NOT use for non-Printify POD vendors, generic Shopify storefront work, or print-file (artwork) generation. Do NOT use for Generate the artwork PNG file that gets uploaded as a print file. Do NOT use for Implement the Shopify side of the Printify-to-Shopify sync. Do NOT use for Design a generic POD-vendor-agnostic product schema.

jacob-balslev Skill

Use when facing an unknown software failure, when symptoms point to different root causes, or when an initial debugging attempt has not converged. Provides a triage-first diagnostic routing framework: classify the failure, collect the right evidence, choose a technique, track confidence, and escalate when stuck. Do NOT use for executing scientific debugging after triage (use `debugging`), code-quality review (use `code-review`), or proactive observability setup. Do NOT use for actually execute scientific-method debugging on this stack trace. Do NOT use for review this AI-generated PR for correctness. Do NOT use for scan this repo for OWASP top 10 vulnerabilities. Do NOT use for design observability instrumentation for this service. Do NOT use for decide which agent should pick up this ticket. Do NOT use for what's the right test pyramid for this feature.

jacob-balslev Skill

Use when a goal, plan, design, strategy, or assumption should be examined by reversing it: ask how to guarantee failure, how the opposite could be true, or what would have to happen for the desired outcome not to occur. Covers objective reversal, failure-path enumeration, obstacle-first planning, opposite-hypothesis testing, and prevention design. Do NOT use for decomposing a problem to irreducible premises (use first-principles-thinking), post-failure root-cause analysis (use debugging), recurring failure clustering (use pattern-recognition), or claim/source grounding (use epistemic-grounding). Do NOT use for Break this pricing model down to irreducible assumptions. Do NOT use for This incident already happened; find the root cause. Do NOT use for Score every webhook failure mode with severity, occurrence, and detectability. Do NOT use for Facilitate a team pre-mortem session with independent written generation.

delorenj Bundle

Consult before ANY coding task. Chooses the optimal coding agent strategy based on task complexity, available free tokens, parallelism potential, and provider quotas. Covers: OpenClaw sub-agents, Codex CLI, Claude Flow swarms/hive-minds, Jules (Google), GitHub Copilot coding models, Augment Code, Kimi K2.5, and OpenAI gpt-5.3-codex. All agents must read this before writing code.

delorenj Bundle

Provision any repo with a stack-agnostic semantic-versioning workflow on mise tasks. `init` installs `.mise/scripts/versioning.sh` and wires `version`, `version:bump[-patch|-minor|-major]`, `version:check`, `version:sync`. Discovers every version-bearing file (package.json, pyproject.toml, Cargo.toml, *.csproj, build.gradle, VERSION, git tags) and keeps them in parity — highest semver wins. Wraps build tasks to bump-patch first, and flags hardcoded self-reported version literals (commander `.version("1.0.0")`, `__version__`, MCP server info) so `--version` can't lie after bumps — derive from the manifest at runtime instead. Use when the user says "add versioning", "set up version bumping", "init versioning", "version this repo", "mise version tasks", "semver workflow", "keep versions in sync", "bump the version across all files", or "--version reports the wrong/stale version". Do NOT use for changelogs/release notes (use product-changelog), publishing to npm/PyPI/crates, or unrelated mise config.

kevinnft Bundle

DSPy: declarative LM programs, auto-optimize prompts, RAG.

kevinnft Skill

Pitfalls, examples, and edge cases for Hermes Kanban workers. The lifecycle itself is auto-injected into every worker's system prompt as KANBAN_GUIDANCE (from agent/prompt_builder.py); this skill is what you load when you want deeper detail on specific scenarios.

kevinnft Bundle

vLLM: high-throughput LLM serving, OpenAI API, quantization.

jacob-balslev Skill

Use when scoring a completed agent task, implementation, document, skill upgrade, or other deliverable against the original request, acceptance criteria, verification evidence, quality rubric, and residual risks before calling it done. Covers skeptical critic review, 1-5 scoring, score ceilings, evidence sufficiency, finding/action capture, and the evaluation-revision loop. Do NOT use for designing eval datasets or graders (use eval-driven-development), line-by-line diff review (use code-review), choosing test levels (use testing-strategy), or designing the overall process and gates before work starts (use methodology). Do NOT use for design a new eval dataset, grader, and hard negatives for this router. Do NOT use for review this pull request line by line for bugs and security issues. Do NOT use for choose unit versus integration versus end-to-end tests for this feature. Do NOT use for design the whole implementation methodology and quality gate sequence before work starts.

jacob-balslev Skill

Use when deciding whether to route a task to Google's Gemini Pro frontier model (current Gemini 3.1 Pro generation, moving to Gemini 3.5 Pro) instead of Claude Opus or GPT-5 — especially for very-large-context reasoning (1M-token input), whole-codebase or long-document analysis, native multimodal input (audio/video/PDF/image in one call), or capability-per-dollar at the frontier. Covers the context window, the 200K context-tier pricing cliff, multimodal support, and the per-lane comparison against Claude Opus / GPT-5. Do NOT use for choosing the cheap/fast tier (use `gemini-flash`), for general agent-system architecture (use `agent-engineering`), or for dispatching among local skills (use `skill-router`). Do NOT use for I just need a cheap classifier for 10k rows. Do NOT use for design the multi-agent orchestration for this system. Do NOT use for which of my local skills should handle this request?

jacob-balslev Skill

Use when deciding whether to run a task in the Claude Code agent harness, when scoping work to its native capabilities (skills, hooks, subagents, MCP, plan mode, 1M-context Opus, background tasks, slash commands), or when choosing Claude Code versus a different harness (Codex, OpenCode, Copilot) for a given piece of work. Covers what the harness is good at, the extensibility stack and when each layer earns its keep, and the decision boundaries against rival harnesses. Do NOT use for writing Anthropic SDK / API code (use `claude-api`), for choosing which Claude MODEL to route a task to versus GPT (use `gpt-5-5`), or for designing a generic autonomous agent loop (use `autonomous-loop-patterns`). Do NOT use for write a Python script that calls the Anthropic Messages API. Do NOT use for is GPT-5.5 or Opus better for this code review? Do NOT use for design a resumable autonomous loop with a supervisor.

jacob-balslev Skill

Use when deciding whether to route a task to Anthropic's frontier reasoning tier (Claude Opus) — architecture, multi-file synthesis, hard debugging, security reasoning, long-horizon agentic planning — and when NOT to (mechanical work belongs on a cheaper tier). Covers the Opus capability profile: 1M-token context, the cost/latency premium, adaptive thinking, the Opus-only effort ceiling (xhigh/max), task budgets, prompt-caching minimums, and high-resolution vision. Do NOT use for picking the balanced implementation tier (use claude-sonnet), the fast/cheap tier (use claude-haiku), designing the loop the model runs inside (use autonomous-loop-patterns), or for Claude API request syntax (read the claude-api reference).

jacob-balslev Skill

Use when reviewing a pull request, diff, or proposed code change for correctness, clarity, security, performance, maintainability, test evidence, and project-convention fit, whether the author is a human, an AI agent, or a peer. Covers pre-review fact gathering, verifying AI-written PR summaries against the diff, reading tests before implementation, tracing call sites and blast radius, review size and attention budget, severity grading with Conventional Comments, comment phrasing, reviewer qualification, treating diff content as evidence rather than instructions, refusing rubber-stamp approval for AI-generated diffs, and making an explicit approve/request-changes/close merge decision. Do NOT use for authoring the code (use refactor for behavior-preserving changes or skill-scaffold for new skills), chasing a known bug after merge (use debugging), security-only audits (use owasp-security), or explaining a patch without a merge verdict (use diff-analysis).

jacob-balslev Skill

This skill provides expertise in data and context compression: SaaS payload optimization (Zstd, Brotli, Gzip), database storage compression, and AI context window compression (Semantic Summarization, Token Pruning). Use when optimizing API latency, reducing storage costs, or managing long-running agent sessions near context limits. Do NOT use for image/video lossy compression (use product-photo) or file archiving.

jacob-balslev Skill

Use when choosing semantically precise names for files/functions/variables/types/columns, resolving overloaded terms, reviewing error messages or UI copy for blame/register clarity, or adapting language for end-user/agent/developer/global-audience contexts. Covers morphology, compound-word order, abbreviation policy, verb-noun naming, polysemy qualification, audience register, blame-free error structure, and cross-cultural language awareness. Do NOT use for casing convention policy (use `naming-conventions`), call-site-wide renames (use `refactor`), docs/navigation structure (use `information-architecture`), specialized UI text pattern catalogs (use `microcopy`), or final prose humanization (use `writing-humanizer`). Do NOT use for decide kebab-case vs snake_case vs camelCase for new database columns. Do NOT use for restructure this doc into a tutorial format with progressive disclosure. Do NOT use for implement Intl.NumberFormat for DKK vs USD currency formatting.

jacob-balslev Skill

Use when diagnosing organizational alignment with the McKinsey 7S framework: shared values, strategy, structure, systems, style, staff, and skills; how the seven elements reinforce or conflict; change-readiness gaps; and implementation risks. Covers internal organization effectiveness, operating-model alignment, transformation diagnostics, post-merger integration checks, and strategy-to-execution gap analysis. Do NOT use for external industry profit-pressure diagnosis (use porters-five-forces), activity-level value and cost mapping (use value-chain-analysis), durable resource advantage testing (use vrio), broad internal/external option generation (use swot-tows), or execution goal-setting alone (use okrs). Do NOT use for Analyze buyer power, supplier power, substitutes, entrants, and rivalry in this industry. Do NOT use for Map the value chain activities that create customer value and drive cost. Do NOT use for Use VRIO to test whether our data, brand, and process are durable advantages.

jacob-balslev Skill

Use when serializing merges across multiple agent branches, resolving conflicts between agent outputs, or cleaning stale task branches. Covers atomic locking, idempotency checks, non-fast-forward handling, and worktree cleanup. Do NOT use for ordinary git operations outside an agent merge queue (use `version-control`).

jacob-balslev Skill

Use when planning multi-step implementations, designing quality gates, establishing verification protocols, or building agent checklists calibrated to known failure modes. Covers methodology/method/process distinctions, Cleanroom, PSP/TSP, hypothesis-driven development, DMAIC, checklist design, V&V frameworks, EDDOps, quality gates, and PDCA. Do NOT use for code-review verdicts (use `code-review`), behavior-preserving implementation work (use `refactor`), or test strategy (use `testing-strategy`). Do NOT use for review this PR and decide whether to approve it. Do NOT use for refactor this file while preserving behavior. Do NOT use for decide unit vs integration vs e2e coverage for this feature. Do NOT use for write the eval cases and grader rubric for this router. Do NOT use for block this dangerous git command or secret-bearing tool call.