More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+22 more

    espalier init

    1

    Discover codebase patterns and generate structured rules, skills, and quality gates for AI coding agents.

    Free

    Security scanned
    espalier init

    espalier init

    Example session with this skill installed

    Set up Espalier for this project. I want it to scan my TypeScript repo and wire it for Claude Code and Copilot.

    • Read your context and instructions
    • Compiled the espalier init

    I have analyzed the repository and completed the Espalier setup process. This included scanning 8 source files to identify your MVC architecture, async/await patterns, and Vitest testing conventions.

    Espalier Initialization Result

    PhaseStatusArtifacts Generated
    DiscoverySuccessMapped 4 layers: controllers, services, models, middleware
    RulesSuccessGenerated coding-standards.md, engineering-structure.md, security-standards.md
    SkillsSuccessCreated espalier-coding, espalier-review, and espalier-testing
    WiringSuccessLinked .claude/rules, .github/skills, and updated CLAUDE.md
    HooksSuccessInstalled pre-push-gate.sh with npm run build && npm test

    Wired: skills=12 rules=5 agents=3

    Next steps

    1. Review espalier/rules/coding-standards.md to ensure the discovered naming patterns match your preferences.
    2. Run /espalier "Add a new endpoint to the users service" to test the coding pipeline.
    3. Reload VS Code to enable the new GitHub Copilot Agent Skills.

    Espalier installed. Wired 12 skills, 5 rules, 3 agents.

    If this saved you time, a ⭐ helps the next person find it:
    https://github.com/Junhanliu-dev/espalier-engineering

    Hit a snag or have a suggestion? An issue is the most useful thing you can send:
    https://github.com/Junhanliu-dev/espalier-engineering/issues/new

    Either way — thanks for trying it.

    Copilot wiring is in place. Notes:

    • Reload VS Code (or restart Copilot CLI) so the Agent Skills register; invoke with /espalier <requirement>.
    • The quality gates (.github/hooks/espalier-gates.json) run in Copilot CLI and the cloud coding agent; VS Code chat does not execute hooks — the pipeline's in-skill gates still apply there.
    • Sub-agents are @harness-coder / @harness-reviewer / @harness-security (.github/agents/).

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Setting up AI agents to work on an existing codebase often leads to "hallucinated" patterns that don't match your project's actual conventions. This skill solves that by automating the discovery of your architecture, naming conventions, and unwritten rules, then encoding them into machine-enforceable constraints.

    What it does

    • Codebase discovery scans your repository to extract real-world architecture, naming patterns, and error-handling strategies.
    • Constraint generation produces structured rules, skill definitions, and wiki pages that guide AI behavior.
    • Quality gate installation sets up pre-push and post-tool hooks to block code that violates your project standards.
    • Multi-platform wiring configures Claude Code, Codex, and GitHub Copilot with project-specific instructions and symlinks.
    • Sub-agent orchestration defines specialized coder, reviewer, and security personas with restricted tool access.

    How it works

    1. Decision phase prompts for merge strategies, tool access scopes, and target platforms.
    2. Scout phase runs parallel analysis tools to map architecture, coding patterns, and security surfaces.
    3. Synthesis phase generates the espalier/ directory containing rules, skills, and project wiki.
    4. Bootstrap phase executes a shell-based wiring process to link the generated system into your IDE and CLI tools.

    Frameworks & tools

    This skill is framework-agnostic but includes specialized scouts for TypeScript, Python, and Go. It integrates directly with Claude Code, GitHub Copilot, and Codex, utilizing shell scripts for quality gate enforcement.

    Why this beats prompting it yourself

    Generic system prompts can't account for the unique, unwritten rules of your specific codebase. This skill discovers those rules from your actual source code, ensuring AI-generated changes follow your exact patterns without manual prompt engineering.

    Use cases

    • Onboarding AI agents to a legacy codebase with complex, undocumented patterns.
    • Standardizing AI output across a large team to ensure consistent architecture.
    • Automating architectural linting and security boundary checks at the agent level.
    • Scaffolding a new project with "Decide-Then-Bind" path to enforce architectural decisions from day one.

    Known limitations

    Requires a functional Bash environment for the bootstrap phase. Language-specific boundary checks are currently optimized for TypeScript, Python, and Go.

    How to use

    Skills on Agensi run inside your AI agent. Copy the prompt, paste it into your agent, and it connects to Agensi and installs this skill for you. If a payment is needed, your agent shows it to you first.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill passes our trust checks below.

    Security scanned

    Passed our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free forever

    No account required to browse

    Trust & safety

    Security scanned

    Verified clean 15 days ago

    • Free to download with an account

    Requires

    Clickup
    Harvestapp
    macOS
    jest
    Node.js
    vitest
    mocha

    Listed15 days ago

    What's inside

    Frequently Asked Questions