afonsoft
UserA curated collection of high-performance Agent Skills (SKILL.md) for Claude Code, Codex, Cursor, OpenCode, and Devin. Covers MCP server building, code review, draw.io architecture, Composio, NotebookLM, SonarQube, and more.
Categories
Indexed Skills (21)
composio-mcp
Use when the user wants to connect AI agents to external apps (Gmail, GitHub, Slack, Notion, Linear, Jira, 1000+ toolkits) via Composio, when `composio` CLI commands fail or need auth, when configuring the Composio MCP server (https://connect.composio.dev/mcp), when `devin mcp list` shows composio failing to list tools, or when the user asks to install/authenticate/use Composio. Covers CLI login, consumer key (ck_*) vs project API key (ak_*), MCP header config, tool search/execute/link, and headless auth flows. Do NOT use for building a new MCP server (use building-mcp-servers). Part of the afonsoft/skills collection.
drawio-architecture
Use when generating architecture diagrams, system/service diagrams, network topology, cloud architecture, flowcharts, ER diagrams, or any draw.io/diagrams.net visualization, AND you need to wire up the official draw.io MCP server (npx @drawio/mcp) or actually open diagrams through MCP tools (open_drawio_xml, open_drawio_mermaid, open_drawio_csv, search_shapes, list/get/set_page). Covers MCP server configuration across Claude Desktop, Claude Code, VS Code/Copilot, Cursor, OpenCode and other stdio MCP clients, plus native .drawio XML authoring (layered/swimlane/nested-container layouts, cloud icons, edge routing, libavoid/ELK layout passes). Do NOT use for non-diagram code generation or image editing. Part of the afonsoft/skills collection.
notebooklm-mcp
Use when the user wants to use Google NotebookLM (Gemini Notebook) with an AI agent via the `nlm` CLI or the `notebooklm-mcp` MCP server, when `devin mcp list` / `claude mcp list` shows `notebooklm-mcp` failing to list tools, when `nlm login --check` fails with `ClientAuthenticationError`, when authenticating NotebookLM on a headless server without a browser, when extracting Google cookies manually or via an external CDP provider (OpenClaw), or when the user asks to configure/authenticate/use NotebookLM MCP. Covers cookie-based auth (manual file mode + OpenClaw CDP), MCP server setup across platforms, and the full `nlm` CLI command surface. Do NOT use for building a new MCP server (use building-mcp-servers). Part of the afonsoft/skills collection.
wordpress-mcp
Use when the user wants to manage WordPress sites via MCP (Model Context Protocol). Covers three paths — (A) the official WordPress/mcp-adapter plugin (Abilities API, 3 meta-tools, HTTP+STDIO transport, v0.6.1+ no composer needed), (B) the AI Engine plugin (43-109+ admin tools: posts, users, media, plugins, options, SEO, social, AI image gen, Gutenberg blocks), and (C) wp-mcp-ultimate (community, 58 abilities, OAuth 2.1, self-contained, WP 6.7+). Includes plugin install via WP-CLI, Application Password / Bearer Token / OAuth setup, STDIO + HTTP transport, per-platform MCP config (Claude Code, Devin, OpenCode, Gemini, Codex, AGY, OpenClaw), endpoint verification, and troubleshooting. Trigger when the user says "WordPress MCP", "mcp-adapter", "AI Engine MCP", "wp-mcp-ultimate", "configure WordPress for agents", or wants to expose WordPress to AI agents over MCP. Do NOT use for building a new MCP server (use building-mcp-servers). Part of the afonsoft/skills collection.
building-mcp-servers
Use when building MCP (Model Context Protocol) servers that let LLMs call external APIs or services, in TypeScript (MCP SDK), Python (FastMCP), or C# (ModelContextProtocol.AspNetCore). Covers tool design, transports (Streamable HTTP/stdio), OAuth 2.1 auth, and evaluations. Do NOT use for consuming or configuring an existing MCP server, or for non-MCP API integrations. Part of the afonsoft/skills collection.
code-review-and-quality
Use when reviewing code before it merges — your own, another agent's, or a human's. Assesses quality across five axes (correctness, readability, architecture, security, performance) before a change enters the main branch. Do NOT use for writing code, or as a substitute for automated tests/CI; this is a review gate, not an implementation guide. Part of the afonsoft/skills collection.
create-readme
Use when generating or updating professional README.md and CHANGELOG.md files for a project. Covers repository analysis, stack detection, badge generation, structure diagrams, and alignment with Keep a Changelog and SemVer standards. Do NOT use for API-reference docs (use dedicated doc tooling), licenses/governance documents, or internal code documentation (use docstrings/JSDoc). Part of the afonsoft/skills collection.
obsidian
Use when working with Obsidian vaults — running the Obsidian CLI (read/create/search/manage notes, tasks, properties), building Bases (.base views/filters/formulas), writing Obsidian Flavored Markdown (wikilinks, embeds, callouts, properties), or developing/debugging plugins and themes. Do NOT use for generic Markdown editors or non-Obsidian note tools; this skill is Obsidian-specific. Part of the afonsoft/skills collection.
sonarqube-review
Use when fixing SonarQube code quality issues automatically across any language or framework — issue analysis, fix generation, unit tests, and coverage. Supports Community, Enterprise, and custom SonarQube deployments via environment variables. Do NOT use for general code review without SonarQube (use code-review-and-quality), for whole-repo quality interventions without SonarQube (use quality-test-implementation), or for non-SonarQube static analysis tools. Part of the afonsoft/skills collection.
create-issues
Use when turning plans, specs, PRDs, and Epics into trackable GitHub Issues. Creates stable Epic IDs, direct links, vertical slices, dependencies, and verifiable acceptance criteria using the GitHub CLI (gh). Part of the afonsoft/skills collection.
diagnose
Disciplined diagnosis and re-validation loop for hard bugs and performance regressions. Reproduce, minimise, hypothesise, instrument, fix, and regression-test. Use when the user says diagnose this / debug this, reports a bug, says something is broken/throwing/failing, or describes a performance regression. User-facing questions and findings must be in Portuguese (pt-BR). Part of the afonsoft/skills collection.
grill-me-with-spec
Use when the user needs to create or refine a feature SPEC SDD before implementation. Interviews the user in Portuguese, builds a design tree, and writes `.specs/SPEC-{YYYYMMDD}-{feature}.md` following the SDD template. Do NOT use for writing implementation code or after the SPEC is approved. Part of the afonsoft/skills collection.
improve-codebase-architecture
Find deepening opportunities in a codebase, informed by the domain language in .claude/CONTEXT.md and the decisions in .specs/SPEC-*.md. Use when the user wants to improve architecture, find refactoring opportunities, consolidate tightly-coupled modules, or make a codebase more testable and AI-navigable. User-facing questions and recommendations must be in Portuguese (pt-BR). Part of the afonsoft/skills collection.
orchestrator
Govern agent-driven projects, audit preconditions, create documentation, turn gaps into GitHub Issues, and coordinate execution, tests, and QA in a continuous loop. Use when starting or running a software project with the afonsoft agent harness. User-facing questions and confirmations must be in Portuguese (pt-BR). Part of the afonsoft/skills collection.
qa-analyst
Use when the user asks for QA analysis, requirement review, test planning, test cases, bug reports, root-cause analysis of defects, or mentions QA, quality assurance, testar essa feature, or revisar. Works in a loop: review requirements, plan tests, create cases, execute, report bugs, and re-validate. User-facing questions and clarifications must be in Portuguese (pt-BR). Part of the afonsoft/skills collection.
scaffold-mvp
Use when starting a new project in an empty repository. Initializes an agile, high-productivity .NET/Blazor/Angular MVP stack with clean engineering boundaries, atomic configurations, and system stability without architectural shortcuts. User-facing questions and recommendations must be in Portuguese (pt-BR). Part of the afonsoft/skills collection.
tdd-spec
Test-driven development using the approved SPEC SDD as the source of truth. Reads `.specs/SPEC-{YYYYMMDD}-{feature}.md`, turns requirements and acceptance criteria into red-green-refactor vertical slices. Use when implementing a feature or bugfix after the SPEC is approved. User-facing questions and recommendations must be in Portuguese (pt-BR). Part of the afonsoft/skills collection.
create-agent-harness
Bootstrap or migrate a complete production-ready agent harness in a repository — CLAUDE.md, AGENTS.md (symlink/reference), .claude/ (settings.json, rules, agents review/plan/test, skills, commands, hooks, memory, knowledge, CONTEXT.md, RULES.md, MEMORY.md, TOOLS.md, WORKFLOWS.md, README.md), .devin/config.json, .opencode/, .cursor/, .gemini/, context engineering and memory protocol. Use when initializing AI agent support in a new repo, migrating legacy harness (AGENTS.md, .agents/, .devin/, .cursorrules, .windsurf) to .claude/, configuring permissions and hooks, or when the agent loses context every session. Supports Claude Code, Devin CLI/Desktop, OpenCode, Cursor, Gemini CLI, Antigravity IDE/CLI, and OpenClaw. Do NOT use for building MCP servers (use building-mcp-servers).
observability-and-instrumentation
Use when adding logging, metrics, tracing, or alerting to code that runs in production, or when a production incident couldn't be diagnosed from available data. Makes system behavior visible and diagnosable. Do NOT use for live debugging of an ongoing failure (use a debugging skill), profiling/optimization (use a performance skill), or launch-day monitoring checklists (use a shipping skill). Part of the afonsoft/skills collection.
execute-tdd-spec
Test-driven development using the approved SPEC SDD as the source of truth. Reads `.specs/SPEC-{YYYYMMDD}-{feature}.md`, turns requirements and acceptance criteria into red-green-refactor vertical slices. Use when the user asks to execute, run or implement from an approved SPEC. User-facing questions and recommendations must be in Portuguese (pt-BR). Part of the afonsoft/skills collection.
sonarqube-autofix
Use when analyzing SonarQube issues and creating SPEC SDDs with the proposed fixes. Classifies issues by type (bug, code smell, security) and generates `.specs/SPEC-{YYYYMMDD}-{issue-key}-{type}.md` for `execute-tdd-spec` to implement. Supports Community, Enterprise, and custom SonarQube deployments via environment variables. Do NOT use for general code review without SonarQube (use code-review-and-quality), for whole-repo quality interventions without SonarQube (use quality-test-implementation), or for non-SonarQube static analysis tools. Part of the afonsoft/skills collection.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.