npx agentspec-cli install d8ac3410-ac20-4304-991e-ec7268504b79 --framework ClaudeCode→ .claude/skills/dig/SKILL.md
# Dig ## Overview Run investigation-only mode to understand the situation before making any changes. Read files, gather context, and ask questions, but do not modify anything unless the user explicitly approves the specific change. ## Trigger Phrases Use this skill when the user indicates investigation or debugging intent, including: - "let's investigate" - "we need to figure out" - "we're debuggin" - Any request to investigate, analyze, or understand before changing code or configs ## Operating Rules - Do not edit files, run write operations, or make changes by default. - Read files, inspect code, run read-only commands, and ask clarifying questions. - If a change is needed, explain what you want to change and why, then ask for explicit permission. - Treat permission as single-use: if additional changes are needed later, ask again. - Stay in investigation mode until the user explicitly ends it. ## Investigation Workflow 1. Clarify the goal and constraints (what needs to be understood, scope, deadlines). 2. Gather context by reading relevant files, logs, configs, or outputs. 3. Ask focused questions to fill gaps (repro steps, environment, expected vs actual). 4. Summarize current understanding and list candidate causes or hypotheses. 5. Propose next investigative steps; request permission before any change. ## Exiting Investigation Mode Exit only when the user explicitly says something like "ok, we're done with investigation". If unclear, ask for confirmation before changing anything.
Built on the Agent Skills standard
Apollo - Investigation Scout
You are the **READ-ONLY INVESTIGATOR** (Apollo) called by other agents to explore codebases, search for patterns, and gather evidence. You NEVER edit files or run commands.
by AgentHub Crawler
investigate-first
Diagnose ambiguous failures before editing. Use for unknown causes, intermittent behavior, performance regressions, or investigations needing evidence-ranked hypotheses.
by AgentHub Crawler
hunt
Invoke when debugging any error, crash, unexpected behavior, or failing test. Finds root cause before applying any fix. Not for code review or new features.
by AgentHub Crawler
Sign in to leave a comment.
ask
Answer questions about the codebase without generating or modifying any code. Use when the user wants to understand how something works, find where something is defined, or explore the project structure. Invoked with /ask.
by AgentHub Crawler