mit einem Klick
project-discovery
Project orientation for unfamiliar codebases. Use when entering a new project, exploring unknown repos, or working on shaky assumptions about build, test, lint, or CI setup.
Menü
Project orientation for unfamiliar codebases. Use when entering a new project, exploring unknown repos, or working on shaky assumptions about build, test, lint, or CI setup.
Refresh the plan to focus on the task at hand. Use when context grew, completed steps muddy it, or you want to clear context and continue in auto mode.
Sequential-wave dispatch for WO chains where output of one feeds the next, shared locks, or shared files prevent fan-out. Use when planning dependent multi-WO landings.
Publish npm packages built with Bun: package.json config, CLI tool packaging, provenance signing, release automation. Use when setting up `publishConfig`/`files`/`bin`, packaging a CLI, enabling `--provenance`, or wiring release-please.
Scaffold a new ComfyUI custom-node repo (TypeScript + bun build, CI, release-please, vitest+pytest) consuming @laurigates/comfy-modal-kit. Use when bootstrapping or init-ing a comfyui node pack.
Dispatch contract for spawning parallel agents covering worktree collisions, scope overflow, and silent exits. Use when fanning out concurrent agents or authoring a lead prompt.
Claude Code health check — scans plugins, settings, hooks, MCP, runtime state, permissions, marketplace with optional fixes. Use when checking project health or troubleshooting setup.
| created | "2025-12-16T00:00:00.000Z" |
| modified | "2026-06-05T00:00:00.000Z" |
| reviewed | "2026-06-05T00:00:00.000Z" |
| name | project-discovery |
| description | Project orientation for unfamiliar codebases. Use when entering a new project, exploring unknown repos, or working on shaky assumptions about build, test, lint, or CI setup. |
| user-invocable | false |
| allowed-tools | Bash(bash *), Read, Grep, Glob, TodoWrite |
| Use this skill when... | Use project-continue instead when... |
|---|---|
| Entering an unfamiliar codebase and need orientation on language/tooling | Resuming known work on a familiar project from PRDs and feature tracker |
| Reasoning shows uncertainty phrases ("not sure what this does") | Already have a clear next task and just need to continue executing |
| Onboarding a fresh clone where build/test commands are unknown | Use project-init instead when scaffolding a brand-new project from scratch |
Systematic project orientation to understand codebase state before making changes. Prevents working on incorrect assumptions by establishing clear context about git state, project structure, and development tooling.
Automatic Activation Detection:
Discovery Capabilities:
Output:
This skill automatically activates when Claude's internal reasoning or responses contain uncertainty phrases like:
Rationale: These phrases indicate Claude is working on incomplete context, which can lead to incorrect assumptions, wrong commands, or inappropriate file edits.
Users can explicitly request project discovery with keywords:
Do NOT activate this skill when:
For fast, consistent project orientation, run the bundled discovery script:
bash "${CLAUDE_PLUGIN_ROOT}/skills/project-discovery/scripts/discover.sh"
This replaces the manual 5-step process with a single execution that outputs structured data covering git state, project type, tooling, documentation, and risk assessment.
For the full manual workflow (5 steps with all commands, risk flags, the summary template, error handling for non-git / empty / large-monorepo / missing-docs cases, and the underlying rationale), see REFERENCE.md.
After project discovery, if user asks for deeper investigation:
Explore agentsecurity-audit agentcode-analysis agentProject discovery establishes baseline context; specialized skills handle deep investigation.
git branch --show-current # Current branch
git status --short --branch # Git state summary
git log --oneline -n 10 # Recent commits
git rev-list --count HEAD...@{u} # Commits ahead/behind remote
scripts/discover.sh already detects project type (manifest files), tooling
(npm scripts, Make targets, CI workflows), and documentation. Read those signals
from the script's structured output, or use the Glob/Grep/Read tools
directly rather than hand-coding ls/find/head shells — see
REFERENCE.md for the full manual command set.
See examples.md for complete discovery outputs for:
For detailed command reference and more examples, see discovery-commands.md and examples.md.