Menu
Skills neste repositório
O SkillsMP coletou 606 skills de merceralex397-collab/meta-skill-engineering. Abra uma skill para revisar a origem e os detalhes.
merceralex397-collab/meta-skill-engineeringMostrando 40 de 606 skills coletadas.
Apply the 5-step synoptic cycle for multi-domain panoramic synthesis when a problem spans 3+ domains and cross-domain interactions matter more than depth in any single domain. Produces unified emergent understanding through Clear (meditate), Open (expand…
Idioma do texto original: inglês
Install skills from GitHub repositories into the agent's skills directory. Use when a user explicitly asks to install a skill by name, install from a GitHub repo/path, or list available installable skills. Do NOT use for creating new skills, modifying…
Idioma do texto original: inglês
Create Grafana dashboards as code with template variables, reusable panels, and provisioning configuration. Triggers on: "grafana dashboard", "provision grafana", "dashboard as code", "grafana json", "template variable grafana", "grafana panel", "grafana…
Idioma do texto original: inglês
Reference Hildegard von Bingen's Physica for natural history knowledge — plants, stones, animals, fish, birds, elements, trees, reptiles, and metals with their hot/cold/warm/cool temperament, medicinal uses, and symbolic properties. Triggers on queries about…
Idioma do texto original: inglês
Set up a Blender scene programmatically via Python (bpy) with objects, materials, lighting, camera, and environment configuration. Triggers on: "set up blender scene", "create 3D scene in blender", "bpy scene setup", "programmatic blender scene", "automated…
Idioma do texto original: inglês
Credit risk data cleaning and variable screening for pre-loan modeling. Use when you have raw credit data requiring quality assessment, missing value analysis, or feature selection before building predictive models. Triggers on credit risk data, loan approval…
Idioma do texto original: inglês
Build production-ready Dataverse SDK solutions for complete business use cases. Use when users describe business scenarios requiring end-to-end Dataverse implementations (CRM workflows, document management, data migration, integrations) rather than isolated…
Idioma do texto original: inglês
Convert dense transformer models to Mixture-of-Experts (MoE) architectures via FFN upcycling, expert initialization strategies, and router training with load-balancing loss. Triggers on dense-to-MoE conversion, FFN upcycling, expert routing implementation, or…
Idioma do texto original: inglês
Deploy ML models to production serving infrastructure (MLflow, BentoML, Seldon Core) with REST/gRPC endpoints, autoscaling, monitoring, and A/B testing. Use when deploying trained models for real-time inference, setting up prediction APIs, or implementing…
Idioma do texto original: inglês
Execute .NET Framework to .NET 8+ upgrades for class libraries, APIs, and Azure Functions. Use when modernizing legacy .NET projects, migrating from packages.config to PackageReference, or updating CI/CD pipelines for .NET 8+ compatibility.
Idioma do texto original: inglês
Install skills, agents, and teams from the agent-almanac registry into supported agentic frameworks (Claude Code, Cursor, OpenCode, etc.) using the agent-almanac CLI. Triggers on: "install from agent-almanac", "agent-almanac install", "setup skills from…
Idioma do texto original: inglês
Modify existing PDFs — merge multiple files, split by page ranges, rotate pages, add watermarks, fill form fields, edit metadata, and redact content using PyPDF2, pikepdf, or pdftk. Use when manipulating existing PDF files rather than creating new ones or…
Idioma do texto original: inglês
Capture desktop screenshots (full screen, app window, or pixel region) on macOS, Linux, and Windows. Triggers on: 'take a screenshot', 'capture my screen', 'screenshot of app', 'screen region', 'window capture'.
Idioma do texto original: inglês
Extract tables from PDFs, HTML pages, and images into structured formats. Triggers on: extracting tabular data from PDF reports or scanned documents, converting document tables to CSV/JSON/DataFrames, parsing tables from HTML without API access, handling…
Idioma do texto original: inglês
Scaffold a new agent skill from this template when you need to create reusable automation instructions. Use when: building a new SKILL.md from scratch, standardizing an ad-hoc workflow into a shareable skill, or preparing a skill for contribution to a…
Idioma do texto original: inglês
Audit a SKILL.md for structural anti-patterns that reduce routing accuracy, output quality, or maintainability. Use when a user says "check this skill for anti-patterns", "what's wrong with this skill", or "audit this skill before promotion". Also use for…
Idioma do texto original: inglês
Audit a skill library for duplicates, category drift, and discoverability gaps; maintain the catalog index, metadata, tags, and naming conventions. Use when: "audit the skill library", "clean up overlapping skills", "organize the catalog before release",…
Idioma do texto original: inglês
Build trigger tests and output-format tests for a skill's evals/ directory. Use when "create tests for this skill", "set up evals", "build a test harness", a new skill needs test coverage, or a skill lacks an evals/ directory. Creates trigger-positive.jsonl,…
Idioma do texto original: inglês
Create new agent skills from scratch and iteratively improve them through test-review-improve cycles. Use this for "create a skill for X", "write a skill that handles Y", "I need a new skill to do Z", "turn this workflow into a skill", or when a repeated task…
Idioma do texto original: inglês
Coordinate multi-agent or multi-stage work so each delegate has a bounded scope, clear artifact contract, and explicit handoff point. Use when the task is large enough for parallel or staged delegation, when one agent must gather evidence while another…
Idioma do texto original: inglês
Find external skills from public registries, GitHub repos, and official skill collections, then evaluate them for quality, licensing, and fitness for adoption. Use when looking for existing skills before building from scratch, evaluating external skill…
Idioma do texto original: inglês
Create a new SKILL.md with YAML frontmatter, routing description, and structured body sections from scratch. Use this for "create a skill for X", "write a skill that handles Y", "I need a new skill to do Z", or when a repeated task pattern should become a…
Idioma do texto original: inglês
Coordinate multi-agent or multi-stage work so each delegate has a bounded scope, clear artifact contract, and explicit handoff point. Use when the task is large enough for parallel or staged delegation, when one agent must gather evidence while another…
Idioma do texto original: inglês
Evolve an existing agent definition by refining its persona in-place or creating an advanced variant. Covers assessing the current agent against best practices, gathering evolution requirements, choosing scope (refinement vs. variant), applying changes to…
Idioma do texto original: inglês
Evolve an existing skill by refining its content in-place or creating an advanced variant. Covers assessing the current skill, gathering evolution requirements, choosing scope (refinement vs. variant), applying changes, updating version metadata, and…
Idioma do texto original: inglês
Evolve an existing team composition by refining its structure in-place or creating a specialized variant. Covers assessing the current team against template and coordination patterns, gathering evolution requirements, choosing scope (adjust members, change…
Idioma do texto original: inglês
Bootstrap a first MCP server from zero to working connection. Use when the user says 'get started with MCP', 'create my first MCP server', 'set up MCP', or wants a hello-world MCP experience. Covers project scaffolding, first tool registration, transport…
Idioma do texto original: inglês
Trigger — "agent needs human input", "when should the agent ask the user", "pause for approval", "escalate to human", "resume after human response", "handle user interrupt mid-run". Skip — when the agent has all information needed to proceed, when the task is…
Idioma do texto original: inglês
Implement a JSON-RPC 2.0 A2A server with full task lifecycle management (submitted/working/completed/failed/canceled/input-required), SSE streaming, and push notifications. Use when implementing an agent that participates in multi-agent A2A workflows,…
Idioma do texto original: inglês
Trigger — "gather context for implementation", "load project conventions", "what files do I need", "trace imports for", "map the API surface", "what patterns does this project use". Skip — if the task is writing code (use `implementer-node-agent`),…
Idioma do texto original: inglês
Trigger — "implement across multiple files", "coordinate implementation", "split work across modules", "implement this plan", "orchestrate the changes", "multi-file implementation". Skip — if only a single file needs changes (use `implementer-node-agent`), if…
Idioma do texto original: inglês
Trigger — "implement in TypeScript", "write the JS module", "add this Node endpoint", "fix this TS file", "create a React component", "update the Express route", "npm package changes". Skip — if the task involves multiple modules (use `implementer-hub`), if…
Idioma do texto original: inglês
Monitors autonomous agent tasks for stalls, infinite loops, resource exhaustion, and progress stagnation, then applies bounded recovery actions. Trigger — "agent is stuck", "task seems hung", "detect stall", "watchdog timer", "loop detection", "agent not…
Idioma do texto original: inglês
Monitor, cap, and recover from context accumulation in agentic systems. Covers per-cycle cost tracking, context window auditing, budget caps with enforcement policies, emergency pruning when approaching limits, and progressive disclosure integration to…
Idioma do texto original: inglês
Designs manager/worker agent hierarchies with explicit span-of-control limits, delegation depth, and communication topology for multi-agent systems. Trigger — "how many agents should report to one manager", "design agent hierarchy", "span of control",…
Idioma do texto original: inglês
Diagnoses failures in multi-agent systems by tracing errors across agent boundaries, replaying message sequences, and constructing blame paths to the root cause. Trigger — "agent produced wrong output", "which agent failed", "trace the error", "debug…
Idioma do texto original: inglês
Runs multiple expert agents in parallel on the same problem, normalizes their outputs, detects consensus or disagreement, and synthesizes a final recommendation. Trigger — "get multiple opinions", "panel of experts", "expert consensus", "compare approaches",…
Idioma do texto original: inglês
Enforces safety guardrails for parallel agent work — prevents write conflicts, ensures file-level isolation, and detects interference between concurrent agent lanes. Trigger — "agents writing to same file", "parallel conflict", "write collision", "lane…
Idioma do texto original: inglês
Design and implement RAG retrieval pipelines — choose chunking strategies, select embedding models, configure vector stores, wire hybrid search with BM25+dense, add reranking stages, and pack retrieved context into LLM prompts. Use when building or debugging…
Idioma do texto original: inglês
Systematically collects facts from a repository to support architectural, planning, or review decisions. Trigger — "gather evidence from the repo", "what does this codebase use", "inventory the project", "analyze dependencies", "detect conventions", "what's…
Idioma do texto original: inglês