explore

Coordinate non-destructive exploration across planning commands with guardrails.

Updated Aug 23, 2026

Search Tags:#planning #codebase #subagents #orchestration #openspec #guardrails #explore

Ready to run

npx skills add https://github.com/CuongTran69/tramthientra --skill explore-cuongtran69

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: explore
Source: https://github.com/CuongTran69/tramthientra/tree/main/.augment/skills/explore
Command: npx skills add https://github.com/CuongTran69/tramthientra --skill explore-cuongtran69

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Shared explore/plan mode behavior for all planning commands (feat, fix, chore, refactor, perf, docs, test, ci, docker). Provides the stance, continuous verification, fluid workflow, subagent protocols, OpenSpec awareness, and guardrails.

Core Features & Use Cases

It defines the shared explore mode that allows reading files, tracing logic, and investigating the codebase without writing changes. It also enforces subagent guidelines and orchestrator-provided guardrails.

Quick Start

From the workspace root, run all OpenSpec and Bash commands; do not cd into subdirectories.

Frequently Asked Questions about explore

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I explore a codebase to plan changes without modifying files?▼

To plan codebase changes safely, use a non-destructive explore mode that coordinates subagents to read files and trace logic while enforcing strict guardrails that prevent any code modifications.

What is subagent orchestration for non-destructive codebase exploration?▼

Subagent orchestration for non-destructive codebase exploration coordinates multiple agents to investigate logic across planning tasks, enforcing strict protocols and guardrails that prevent any code modifications.

Can I use explore mode for planning tasks like ci, docker, and test?▼

Yes, explore mode supports planning tasks like ci, docker, and test by applying a domain-agnostic stance and subagent protocols across all these commands to prevent code modifications.

How do I trace logic across a codebase without writing changes?▼

You can trace logic across a codebase without writing changes by running commands from the workspace root in explore mode, which enforces continuous verification and orchestrator-provided guardrails.

Does OpenSpec exploration require running commands from the workspace root?▼

Yes, OpenSpec exploration requires running all commands from the workspace root rather than cd-ing into subdirectories, ensuring the explore mode maintains consistent guardrails across the codebase.