11-explore — Claude Code skill: what it does, how to install
11-explore surveys a project along three axes: installed tooling, context and codebase. You get a compact map, then a deeper walk down whichever axis you name.
MIT ★ 571 updated 2026-10-29
/plugin marketplace add ai-driven-dev/framework
/plugin install aidd-context@aidd-frameworkWhat 11-explore does
The skill covers a project from three sides. Tooling is what is installed: skills, agents, slash commands, rules, hooks, MCP servers and plugins. Context is the memory bank, specs and plans, and the AI context files. Codebase is the stack and the rough layout of the repo. It reports what exists and stops there.
Two actions do the work. survey returns a short map from the project root; drill descends one axis or all of them, level by level, and can aim at a stated goal. Before either runs, the skill works out which AI tools the project uses and takes per-tool scan paths and formats from references/ai-mapping.md. The action's own file in actions/ is read first.
It points, it does not steer: you get an item's invocation path, not a plan. Picking the next move belongs to the onboard skill, so go there for direction. Every level waits for your confirmation, which means one call will not dump the whole tree. Pass the scope as tooling | context | codebase | goal.
What it helps with
- See which skills, hooks and MCP servers a project already carries
- Match a skill or rule to a goal you have in mind
- Get the stack and layout of an unfamiliar repo at a glance
- Check what sits in the memory bank and AI context files
- Walk a single axis down to the level you need
Example prompts for the agent
- Map this project across all three axes
- Drill into the tooling axis and list what's installed
- Which skill fits this goal: adding tests?
How to install 11-explore
1Plugin marketplace
/plugin marketplace add ai-driven-dev/framework/plugin install aidd-context@aidd-frameworkThis installs the whole aidd-context plugin with all its skills
2Skills CLI (Vercel)
npx skills add ai-driven-dev/framework --skill 11-explore -a claude-code3Manually: put the files in place
~/.claude/skills/11-explore/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/11-explore/
1Skills CLI (Vercel)
npx skills add ai-driven-dev/framework --skill 11-explore -a codex2Manually: put the files in place
.agents/skills/11-explore/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add ai-driven-dev/framework --skill 11-explore -a cursor2Manually: put the files in place
.cursor/skills/11-explore/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install ai-driven-dev/framework plugins/aidd-context/skills/11-explore2Manually: put the files in place
.github/skills/11-explore/SKILL.md1Skills CLI (Vercel)
npx skills add ai-driven-dev/framework --skill 11-explore -a gemini-cli2Platform CLI
gemini skills install https://github.com/ai-driven-dev/framework --path plugins/aidd-context/skills/11-exploreManually: put the files in place
.devin/skills/11-explore/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add ai-driven-dev/framework --skill 11-explore -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about 11-explore
How do I install 11-explore in Claude Code?
Use the plugin marketplace: /plugin marketplace add ai-driven-dev/framework, then /plugin install aidd-context@aidd-framework.
What does 11-explore do?
It maps a project across tooling, context and codebase, then goes deeper into the axis you pick. Nothing gets executed; you get the invocation path of whatever it finds.
Will it tell me what to do next?
No. Exploration only shows what is there. Choosing the next step is left to the onboard skill.
Source text
SKILL.md · MIT
--- name: 11-explore description: Explore the current project across its tooling, context, and codebase. Use to survey what is installed, see what is available, or find which skill, agent, or rule fits a goal. Not for choosing the next step or running an item; it only points. argument-hint: tooling | context | codebase | goal --- # Explore Surveys the current project across three axes so the user sees what is there and can dig into any of them. It maps the project, it never prescribes a next step. ## Axes - **Tooling**: the AIDD capabilities installed, the skills, agents, commands, rules, hooks, MCP servers, and plugins. What the user can run. - **Context**: the context layer, the memory bank, the specs and plans, and the AI context files. What the AI knows about the project. - **Codebase**: the project itself, the stack and the high-level structure. What the project is. ## Actions | # | Action | Role | Input | | --- | -------- | ---------------------------------------------------------------- | ----------------- | | 01 | `survey` | Scan the three axes and present a compact map | project root | | 02 | `drill` | Descend one axis or all, level by level, match an intent if any | a scope or a goal | Detect the project's AI tools first, from the signals in `references/ai-mapping.md`. Then route by what the user asked: no scope given runs `survey` for a compact map, a named axis or "all" runs `drill` straight away. Always propose the axes when the request is open, never assume one. Run each action's `## Test` before the next. Before running an action, read its file in `actions/`, not only the table or assets. ## Transversal rules - Map, never prescribe. Explore shows what is there across the axes and descends into it on request. It never tells the user the next step, that is the onboard skill's job, at any depth. - Detect the tools before either action. A surface a tool does not have is skipped, never an error. - Propose, then accept. Offer the axes or all when the request is open, run the user's pick directly when they name one. - Descend one level per confirmation. Wait for the user at every level, stop at a leaf or when they are done. - List only what is actually installed or present, never invent an item. - Never hardcode a tool. Per-tool scan paths and formats live in `references/ai-mapping.md`. - Point, do not run. Return an item's invocation path and stop. ## References - `references/ai-mapping.md`: the per-tool signals, scan paths, and formats for the Tooling and Context axes.
© AI-Driven Dev · MIT · ai-driven-dev/framework