What problem does it solve? Setting up an AI agent orchestration framework in an existing project requires manually detecting the language, test runner, linter, and CI system, then writing platform-specific instruction files and coverage configuration by hand. This Skill automates that entire onboarding flow through interactive detection and guided questions. ## Core Features & Use Cases - Automatic Project Detection: Scans marker files to identify language, framework, package manager, test runner, linter, formatter, type checker, CI, and git hooks without asking the user. - Platform-Specific Configuration: Writes or appends instruction files for Codex (AGENTS.md), Claude (CLAUDE.md), Gemini (GEMINI.md), or OpenCode, plus coverage thresholds and Claude command shims via a mandatory setup script. - Project Profile & Knowledge Base: Creates docs/project-profile.json with all detection results and seeds a JSONL knowledge base, shell utilities, CI templates, and gitignore entries. - Use Case: A developer installs Seal Harness in an existing TypeScript/React repository and invokes the setup skill; it detects vitest, pnpm, ESLint, and GitHub Actions, asks for a coverage threshold, then writes all required configuration files in one pass. ## Quick Start Run the setup skill in your project root and answer the interactive questions to configure Seal Harness for your detected stack.
