mcpField
E
explain

explain

github.com/laststanceDevelopment143 installsUpdated 2w ago
Repository

About explain

A deep explanation skill for experienced developers that analyzes code, traces dependencies, and structures a systematic walkthrough. It emphasizes checking claims against source code and official documentation, and calling out uncertainty.

Key features

  1. 01Inspects target code and call sites
  2. 02Traces dependencies and external APIs
  3. 03Structures explanations by scope or flow
  4. 04Explains architectural and performance choices
  5. 05Validates claims and flags uncertainty

Use cases

  1. 01Understand a function, class, or system
  2. 02Trace dependencies and data flow in code
  3. 03Analyze framework behavior or error messages

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/laststance/skills --skill explain

After installing, type/explainin Claude Code to confirm it loaded.

Details

From the project README

Explain — Deep Systematic Explanation

Codex Compatibility

When running this skill in Codex, translate Claude Code-only primitives before acting: AskUserQuestion -> chat/request_user_input, TodoWrite -> update_plan, Task/TaskCreate/TeamCreate/SendMessage -> spawn_agent/send_input/wait_agent when available and allowed, and EnterPlanMode/ExitPlanMode -> a concise chat plan plus explicit approval. Resolve Read/Write/Edit/Bash/WebSearch/WebFetch to Codex file/shell/web tools, and map ~/.claude/... paths to ~/.agents/... or ~/.codex/... unless the task explicitly targets Claude Code.

Cursor Compatibility

When running this skill in Cursor Agent, translate Claude Code-only primitives before acting: AskUserQuestion -> AskQuestion; TodoWrite -> Cursor TodoWrite or an equivalent checklist; Task/TaskCreate/TeamCreate/SendMessage/multi-agent flows -> Cursor Task (subagents), parallel Tasks, or run_in_background when allowed (TeamCreate/SendMessage may have no exact match); EnterPlanMode/ExitPlanMode -> Plan mode (SwitchMode / CreatePlan) plus explicit user approval. Resolve Read/Write/Edit/StrReplace/Bash/web/search/MCP via Cursor Composer or Agent equivalents. MCP names written as mcp__server__tool typically map to call_mcp_tool with configured server identifiers. Map ~/.claude/... to ~/.cursor/skills/, .cursor/skills/, and .cursor/rules/ unless the task explicitly targets Claude Code.

Advanced-level explanation with introspection markers, systematic analysis, and accuracy validation.

<essential_principles>

Always Active

  • Advanced level: Assume the reader is an experienced developer. No hand-holding. Explain the "why" and architectural reasoning, not just the "what"
  • Introspection markers: Make reasoning visible throughout:
    • 🤔 Reasoning — "🤔 This pattern suggests..."
    • 🎯 Decision — "🎯 The author chose X over Y because..."
    • ⚡ Performance — "⚡ This has O(n²) complexity due to..."
    • 📊 Quality — "📊 This follows/violates the SRP principle"
    • 💡 Insight — "💡 This pattern can be applied to..."
  • Validate before presenting: Never guess. Read the actual code. Check official docs when explaining library/framework behavior. Flag uncertainty explicitly
  • No project-specific rules: This skill works across all projects and AI agents

</essential_principles>

Phase 1: Analyze

Thoroughly examine the target before explaining anything.

  1. Identify scope: Is this a file, function, class, concept, or system?
  2. Read the code: Use Read, Grep, Glob to gather the actual source
  3. Trace dependencies: Follow imports, references, and call sites to understand context
  4. Check external APIs: Use Context7 or WebSearch for library/framework specifics if needed
  5. 🤔 Form a mental model of how the target works

Tools: Read, Grep, Glob, mcp__serena__find_symbol, mcp__serena__get_symbols_overview, mcp__context7__query-docs

Phase 2: Structure

Plan the explanation for maximum clarity.

  1. Identify the core insight — the one thing that unlocks understanding
  2. Decide explanation order:
    • Top-down: Start with the big picture, zoom into details
    • Bottom-up: Start with building blocks, compose into the whole
    • Flow-based: Follow data/execution flow from input to output
  3. 🎯 Choose the approach that best fits the target

Phase 3: Explain

Deliver the explanation with depth and clarity.

  1. Open with context: What is this, and why does it exist?
  2. Core explanation: How it works, with introspection markers throughout
  3. Key decisions: Why was it built this way? What alternatives exist?
  4. Connections: How does it relate to the broader system/pattern?
  5. Code examples: Reference actual line numbers (file:line) when explaining code
Explanation Format

Use the insight block format for key educational points:

`★ Insight ─────────────────────────────────────`
[2-3 key educational points]
`─────────────────────────────────────────────────`

Phase 4: Validate

Verify the explanation is accurate and complete.

  1. ✅ Cross-check claims against actual code (re-read if needed)
  2. ✅ Verify library/framework behavior against official docs
  3. ✅ Flag anything uncertain with "⚠️ Note: ..."
  4. ✅ Ensure no oversimplification that would mislead

Examples

/explain src/auth/middleware.ts
/explain "how does React Suspense work"
/explain the hook system in this project
/explain this error message: "Cannot read property of undefined"

Boundaries

Will:

  • Provide deep, accurate explanations with reasoning made visible
  • Read and trace actual code rather than guessing from filenames
  • Reference official documentation for framework/library specifics
  • Flag uncertainty and knowledge gaps honestly

Will Not:

  • Provide shallow or beginner-level explanations (use Claude directly for that)
  • Guess at behavior without reading the source
  • Modify any code (this is a read-only explanation skill)

Comments

Markdown supported