Progressive-disclosure instructions an agent loads on demand. Cheap until invoked, which is why the per-session figure on each card is usually two digits.
Route research, investigation, deep-dive, and source-filing requests through pensar's package-owned autoresearch workflow. Use when a user asks to research a topic into a pensar vault, compare agent skill systems, file sources, or produce a source-linked wiki synthesis.
Debug shaders live in a running Godot game through the godot-mcp CLI: snapshot shader state, tweak and reset uniforms, visualize UV/normals/positions, capture rendered frames, hot-reload shader source with rollback, measure frame times, and check shader warnings and project health. Use when a shader compiles but…
Generate a compact first-pass repository briefing with context-pack. Use when work starts in an unfamiliar repo, when the user asks for orientation or high-signal files, when active changes need a compact summary, or when prompt budget matters.
Run one iteration of the doc improvement loop: generate or accept questions, answer them in parallel, verify, grade, diagnose, fix the docs, and re-evaluate until convergence.
A development and debugging guide for vLLM and vLLM-Ascend, tools for running machine-learning models to generate responses. It covers inference accuracy checks, service management, automated tests, and code fixes for single-machine or split deployments.
LLM-compiled knowledge base management. Ingests primary sources (papers, notebooks, data), distills them into interlinked org-mode wiki articles, and provides structured search over the wiki graph database. Triggers: "ingest this", "distill", "add to crucible", "crucible ingest", "crucible search", "search the wiki"…
Check each provider for newly released models, compare against the supported set, add any missing ones, and update the per-provider docs and website accordingly.
The crank-loop methodology — a measured, iterative loop for delegating implementation tasks to an agent with specs, gates, telemetry, and scoring. Use when running or improving task-spec-driven delegation workflows.
Helps recruiters identify what makes a role genuinely exciting and rewrite job posts to attract top candidates. Uses "So what?" reframing and software engineer motivation frameworks. Use when the user says "make this role appealing", "improve this job ad", "sell this role", "pitch this vacancy", or wants to create…
A workflow for creating a custom 3D personal website from a person's résumé, experience, and design direction. It can produce styles such as interactive game worlds, driving explorations, and archival interfaces.
A browser-based operations assistant for Xiaohongshu, a Chinese social-media and shopping platform. It uses a logged-in Chrome browser to analyse accounts and content and manage publishing workflows.
Visual design guidance for bold, minimal presentations. Provides layout patterns, typography hierarchy, color specifications, and slide composition rules. Use when asking "how should this slide look?", "design guidance for...", "what layout for this slide?", or when translating content into visual structure for a…
Write correct, idiomatic THRML code, centered on Ising machines and how they are trained. THRML is Extropic's JAX-based GPU simulator for the block Gibbs sampling programs that run natively on Extropic's probabilistic computing hardware. Use this skill whenever the user mentions THRML, Extropic, probabilistic…
Development guidance for 易语言 (Easy Language), a Chinese programming language and its project formats, including .e and .ec files. It covers reading, editing, debugging, compiling, and migrating projects.
An iOS project-structure guide that groups code by app feature, shared infrastructure, and reusable design components. Each feature keeps its screens, view models, services, and views together.
A Chinese-language toolkit for making and checking academic charts, diagrams, flowcharts, and graphical abstracts. A graphical abstract is a visual summary of a research paper, while a flowchart shows the steps or decisions in a process.
Prevents overengineering when designing, implementing, refactoring, or reviewing software. Use when the user asks to implement a feature, add an abstraction, refactor code, review a design, simplify existing code, or make an architectural decision. Also triggers when the user asks if code is too complex, whether to…
Drafts concise release notes from merged pull requests and conventional commits. Use when the user asks for a changelog, release notes, or a GitHub release body.
Pixel-faithful clone of any web UI into the user's existing stack, using whatever sources are available — a screenshot alone, a live URL, raw HTML/CSS, or any combination. Use this skill whenever the user wants to recreate, match, replicate, or "clone" a design from a screenshot, image, URL, Figma export, or HTML…
Turns past decisions into calibrated judgment - scoring old predictions against what happened, assigning credit for delayed results, and separating a bad decision from a good decision that lost. Use during a retrospective, postmortem, quarterly or monthly review, when reviewing decisions taken 30+ days ago, when…
Build or review gRPC services and clients in .NET with correct contract-first design, streaming behavior, transport assumptions, and backend service integration.
A Chinese-language guide for writing and revising artificial-intelligence research papers for academic conferences.
★not rated 12▲
+1 yesterdayA68 tokens
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: