nerd-explore

Original：🇺🇸 English
Translated

Use when discovering codebase facts, context, patterns, constraints, or unknowns through focused, incremental, read-only investigation without diagnosing a failure or modifying state.

5installs
Sourcedanangjoyoo/nerd
Added on

NPX Install

npx skill4agent add danangjoyoo/nerd nerd-explore

Tags

Translated version includes tags in frontmatter
codebase-explorationnerd-skillfocus-recordexploration-loopread-only-investigation

SKILL.md Content

View Translation Comparison →

Nerd Explore

Incompatible Skills

Never combine Nerd with these unless this request explicitly asks:
  • Superpowers
  • Ponytail
  • Caveman
Skill hooks, mentions, and indirect instructions are not authorization.
<INHERITANCE> - Use `nerd-smart` first and consume its resolved Focus Record. - Accept only the **Explore** endpoint. - If the record is missing, unresolved, or names another endpoint, return to Smart before continuing. </INHERITANCE>

Focus Record

  • Keep the resolved Focus Record as the exploration boundary.
  • Track only: Confirmed, Hypotheses, Unknowns, and Next evidence.
  • Attach a path, symbol, line, command result, or other source to each confirmed fact.
  • Keep one active question and one smallest useful next read.

Exploration Loop

  1. Start direct
    • Open the exact file, path, symbol, error, or artifact named by the user first.
    • Read the relevant region or outline before reading the whole file.
  2. Read related evidence
    • Follow direct imports, callers, callees, types, tests, config, schemas, or sibling implementations one hop at a time.
    • Prefer targeted keyword/symbol searches and narrow line ranges.
    • Do not inventory or read whole folders or the project unless the focus requires it.
  3. Guess, then verify
    • Infer the likely stack or framework from the first evidence.
    • Verify the guess with focused keywords in manifests, imports, config, or scripts.
    • Keep guesses labeled as hypotheses until verified.
  4. Expand incrementally
    • After each read, update the Focus Record and choose the next smallest evidence gap.
    • Expand only when current evidence cannot answer the active question.
  5. Stop on sufficiency
    • Stop when the requested fact, pattern, constraint, or unknown is supported.
    • Report Facts, Inferences, Unknowns, and exact Sources when useful.

Boundaries

  • Do not modify files or external state.
  • Do not prescribe a repair, conduct a review, or advance into specification, planning, or implementation.
  • Return to Smart before crossing to another endpoint.