Designing agent harnesses and tool systems — risk taxonomy for tools, permission decisions, draft/commit pattern, structured tool results, agent budgets (10 types), context trust labels against prompt injection, plan-artifact, approval records, observability and traces, evals (13 categories), event model, streaming buffering, 3rd-party skill install checklist, agentic RAG, self-improving SOP loops, model policy, reasoning effort, and Programmatic Tool Calling adoption gates. Use when building a new Agent SDK app, custom orchestrator, MCP server, Cloudflare Worker with tool calls, agentic RAG pipeline, model router, or model-tier policy; when designing tools and permissions; when writing an agent loop; or when you need trust labels for external content. Do NOT use for improving or auditing an already-built harness (use harness-audit / harness-design instead), nor for ordinary Claude Code sessions where the harness is already given.
2132d agoReviewMountCreate a free account to mount
Use when designing, auditing, refactoring, or explaining an agentic harness for any domain, especially when work must continue from a measured gap to verified completion. Covers provider-neutral loops, tools, permissions, environment adaptation, planning, durable workflow state, context, skills, observability, evals, and safety for OpenAI, Anthropic, and compatible APIs. Not for implementing an ordinary app feature or reviewing one concrete diff.
2132d agoReviewMountCreate a free account to mount
Практическая инженерия диффузионных моделей: архитектуры, обучение, инференс, оптимизация памяти. Использовать при любых задачах с диффузионными моделями: проектирование или модификация архитектуры (UNet/DiT/Flow/Flux), выбор и настройка schedulers/samplers, дообучение (LoRA/DreamBooth/full fine-tune), оптимизация памяти (AMP/checkpointing/ZeRO/FSDP/quantization), замена или fusion текст-энкодеров (CLIP/Qwen), работа с Diffusers, отладка диффузионных пайплайнов, оценка качества (FID/CLIPScore/LPIPS), latent diffusion, VAE, guidance/CFG, rectified flow, Stable Diffusion, SDXL, Flux. Также применять при вопросах про GPU-память при обучении генеративных моделей, text-to-image пайплайны, ControlNet, multi-encoder fusion, WebDataset. Do NOT use for writing FLUX.2 Klein generation prompts (use flux2-klein-prompting) или для FLUX.2 Klein / Qwen-Edit LoRA-тренинга и edit-LoRA (use flux2-lora-training); это общая диффузионная инженерия, не специфика FLUX.2 Klein.
2132d agoReviewMountCreate a free account to mount
Plan or review LoRA and edit-training work specifically for FLUX.2 Klein or Qwen-Image-Edit, including paired datasets, trainer-version contracts, and held-out fidelity checks. Do not use for generic Stable Diffusion/DiT training, prompt authoring, or model serving; route those tasks to their specialized skill.
2132d agoReviewMountCreate a free account to mount
Forensic image-to-prompt compiler for image generation models. Use this skill whenever the user wants to: convert/describe an existing image into a generation prompt, reconstruct a scene as a prompt, generate prompts from reference images for AI image tools (Midjourney, FLUX, Stable Diffusion, DALL-E, or any diffusion model), write prompts that preserve exact visual properties of a source image, or needs precise control over identity-safe subject description, geometry lock, lighting reconstruction, color anchoring, or handler-based special cases (floating scenes, collages, close-ups, jewelry, garments, surreal elements). Also trigger for requests involving: image editing prompts, reference-driven generation, pose description, camera angle locking, fabric/material description, or any "turn this image into a prompt" task. Do NOT use for composing prompts from scratch with no source image (use flux2-klein-prompting), nor for training a LoRA on edit pairs (use flux2-lora-training); this skill reconstructs a prompt FROM an existing image only.
2132d agoReviewMountCreate a free account to mount
Machine-learning research loop for dataset curation, fine-tuning, evaluation, inference deployment, experiment tracking, and model explainability. Use when working on ML experiments, training data, model benchmarks, RunPod/GPU runs, classifier quality, vLLM/GGUF serving, SHAP-style model explanations, or research-to-code iterations. Do not use for a simple code edit that has no ML dataset, metric, model, or experiment artifact.
2132d agoReviewMountCreate a free account to mount
Use when: NotebookLM, notebooklm MCP, large documentation sets, courses, books, papers, or citation-backed research are mentioned. Retrieves a small grounded answer from a stable corpus, preserves citations, and verifies claims against primary documentation, repository code, and tests. Do not use when: the answer is already in a small local file, the source is rapidly changing, or a live runtime/test is the authority.
2132d agoReviewMountCreate a free account to mount
Choose and evaluate VLM or segmentation pipelines, including text-conditioned detection, masks, part labels, model-license constraints, and measured GPU deployment choices. Use when a task has a VLM or segmentation component; route pure diffusion prompting, training, or serving to its specialized skill.
2132d agoReviewMountCreate a free account to mount
Scaffold a feature narrative in an existing project layer with Design, Plan, Verify, and Conclusion sections. Use when: "create a new feature", "start work on feature", "scaffold feature doc", "/feature-new", or "begin feature narrative". Inspects and preserves the target project's feature ID, registry, filename, and coordination conventions; does not create a missing layer (use /layer-new).
2132d agoReviewMountCreate a free account to mount
Design a scoped multi-agent harness for a requested long-running AI workflow. Preserve the user-approved product boundary while using Generator-Evaluator separation, testable sprint contracts, context management, and task-appropriate validation. Based on Anthropic Engineering patterns. Use when: "build a harness", "multi-agent architecture", "agent orchestration", "generator-evaluator", "long-running app", "harness design", "agent pipeline", "quality evaluation loop", "sprint contract", "build app with agents", "Claude Agent SDK architecture", or when building complex full-stack apps that need planning → generation → evaluation cycles. Also use when discussing context degradation, self-evaluation bias, or assumption testing in AI workflows. Do NOT use to stress-test or critique an already-written plan document; use plan-swarm-review for that (this skill designs the harness, it does not review plans).
2132d agoReviewMountCreate a free account to mount
Scaffold a new layer in a project's docs/layers/ tree following the feature-layer architecture (principle 28). A layer is a bounded concern (security, data, ui, infrastructure, domain) with its own invariants, decisions, gotchas, patterns, and feature narratives. Use when: "create a new layer", "add security layer", "scaffold layer", "start tracking <concern> separately", "/layer-new", "add bounded concern". Operates on the kb-skeleton structure; idempotent -- will not overwrite existing layers. Do NOT use to scaffold an individual feature narrative inside an existing layer; use /feature-new for that (a layer is the container, not the per-feature doc).
2132d agoReviewMountCreate a free account to mount
Iterative plan review using multisampling + focused decomposition. Launches parallel independent agents to find issues that single-pass review misses. Up to four bounded rounds: broad, multisample, focused, focused-plus-multisample. Use when: "swarm review", "review plan thoroughly", "multisample review", "deep plan review", "plan swarming", "stress test the plan", or when evidenced cross-component risk warrants deeper plan review. Size alone does not require a swarm or another review round. Do NOT use to design a multi-agent harness or Generator-Evaluator architecture from scratch; use harness-design for that. Ordinary code/diff review routes to deep-review; use this skill's code mode only when the user explicitly requests a swarm security/bug review of code.
2132d agoReviewMountCreate a free account to mount
Convert short scene descriptions, book/album cover briefs, or 2-paragraph synopses into seamless-loop animated pixel-art covers rendered as self-contained HTML+canvas. Companion skill to pixel-art-studio. Use when the user asks to "make a cover for", "animated book cover", "looped pixel scene", "convert this story description to pixel art", "create ambient pixel animation", "обложка для книги в пиксель-арте", "анимированная обложка", "封面 像素画", "픽셀 아트 표지", or provides a short narrative/synopsis and wants a visual result. Covers: 5-element scene framework (Subject + Setting + Lighting + Palette + Motion), iconographic shorthand for symbolic accents, seamless loop techniques (phase-based parametric, sub-pixel breathing, LCM-clean parallax, deterministic particle systems), loop period selection by mood, three prompt registers (LLM agent / human artist / SDXL LoRA), single-HTML-file deliverable in dark-atmospheric style with parametrized canvas rendering. Generates the same engine pattern as the user's `Grass Field with City.html` reference (canvas + requestAnimationFrame + phase-derived parameters). Do NOT use for individual game sprites, character sheets, walk/attack cycles, or PNG/GIF sprite-sheet export; use pixel-art-studio for that (this skill makes looped HTML+canvas scene covers, not raster sprite assets).
2132d agoReviewMountCreate a free account to mount
Decide module boundaries before the first file: what modules exist, which way dependencies point, who owns state, and what each module may know. Use when starting a project, service, site, API, or subsystem; adding a feature with no obvious home; resolving a circular import or inverted framework dependency; or writing an ARCHITECTURE.md or ADR. Do not use for a one-file script, throwaway experiment, bug fix inside an established seam, naming/function-shape cleanup (use code-complexity), an existing oversized module (use refactoring-safely), or capacity/data scaling decisions (use system-and-data-design). This defines earned boundaries; it does not license speculative layers.
2132d agoReviewMountCreate a free account to mount
Keep web applications, APIs and services readable as they grow: choose feature or domain seams, assign state ownership, enforce dependency direction, keep adapters thin, and verify file shape. Use when starting or extending a web app, backend, frontend, API or multi-page product; when a change makes a module hard to read; when architecture review finds a god file, cross-feature imports, a circular dependency or a framework-heavy domain. Load architecture-first first for a new system, and refactoring-safely for an existing oversized module. Do not use for a one-file script, throwaway spike or a purely local naming change.
2132d agoReviewMountCreate a free account to mount
Find likely software bugs in a codebase, rank concrete bug candidates, and prove or reject them with focused regression tests before proposing a fix. Also turn bug reports, stack traces, screenshots, failing behavior, support tickets, and regressions into minimal reproducible cases with red-to-green evidence. Use when Codex needs to hunt for unknown bugs, audit code for correctness defects, test suspicious edge cases, reproduce a reported failure, isolate root cause, or verify that an approved fix works without regressions. Honor explicit user authority for the requested fix; use approval gates only for authority that the request did not already grant or when scope materially changes. Do NOT use for ordinary implementation where no bug investigation or reproduction is needed.
2132d agoReviewMountCreate a free account to mount
Keep each unit comprehensible while the code grows: deep modules over shallow ones, information hiding, honest names, small functions with one job, error handling that does not lose the error, DRY and orthogonality, design by contract, and not leaving broken windows. Merges module-complexity management with naming/function/error-handling discipline and the pragmatic meta-rules. Use when writing or reviewing a function, class or module; when an interface feels wide, a class feels shallow, or a change in one place forces edits in three; when the user says "this is hard to follow", "clean this up", "review this code", "is this a good abstraction", "too many parameters", "pass-through method", "duplicated logic", or asks about naming, comments, error handling or unit tests. Do NOT use to decide where a new module LIVES or which modules exist (use architecture-first), to plan capacity or storage (use system-and-data-design), to execute a named transformation on a file that is already too large (use refactoring-safely), or to strip over-engineering on request (use lean-code). This makes a unit comprehensible; it does not draw the system's boundaries.
2132d agoReviewMountCreate a free account to mount
Drive and inspect an interactive CLI or TUI with a repeatable local harness, deterministic input, transcripts, and optional profiling. Use for CLI UX checks, prompt flows, startup regressions, hangs, interrupts, resize behavior, or memory growth. Do not use for a non-interactive command that a normal test can cover.
2132d agoReviewMountCreate a free account to mount
Drive and inspect a local web, IDE, or Electron UI with browser or CDP automation and evidence. Use for UI regressions, screenshots, accessibility snapshots, visual diffs, focus, keyboard, scrolling, resize, console, network, performance, or heap checks. Do not use for backend-only changes or for a screenshot without a behavior claim.
A folder with a SKILL.md file: a name, a description of when to use it, and instructions. Claude loads a skill only when the task matches its description.
How do I use one I find here?
Copy the folder into your project's .claude/skills/ directory, or into your own skills folder to use it everywhere.
What do the warnings mean?
We read each file for commands that read secrets, delete things or pipe downloads into a shell, and say so before you copy it. No warning is not a promise that a file is safe.
Which skills worked for people?
Open a skill to see its discussion. Reports from people and their agents are coming.
AnastasiyaWSkill