Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    hypr repo guide

    1

    Generate evidence-backed repository maps and Q&A packs with file-and-line citations and HTML exports.

    Free

    Security scanned
    hypr repo guide

    hypr repo guide

    Example session with this skill installed

    Explain how the authentication flow works in this repository and generate an answer pack.

    • Read your context and instructions
    • Compiled the hypr repo guide
    • Generated the document

    I have inspected the repository's authentication logic, focusing on middleware and provider configurations. I've captured the relevant source passages and validated the citations against the current codebase.

    Authentication Flow Analysis

    The authentication flow is primarily handled through a centralized middleware pattern as seen in the captured evidence.

    ComponentPathResponsibility
    Middlewaresrc/auth/middleware.pyValidates JWT tokens on incoming requests.
    Providersrc/auth/providers/oauth.pyInterfaces with external identity providers.
    Sessionsrc/auth/session.pyManages local session state and cookie rotation.

    Summary of Findings

    Token Validation: The system expects a Bearer token in the Authorization header. This is enforced in middleware.py:45-60.
    2. Provider Logic: The OAuth provider handles the handshake and user profile mapping in oauth.py:112-140.
    3. Documented Setup: The command npm run auth:init is documented in README.md for initializing local secrets.

    Deliverables:

    • answer.md - Full technical explanation with citations.
    • answer.html - Printable version of the analysis.
    • evidence.json - Retained source passages.
    • checks.json - Validation logs.

    Next steps

    1. Review the Unknowns section in answer.md regarding the missing JWT_SECRET configuration.
    2. Open answer.html for a formatted view of the code citations.
    3. Use the evidence.json file to audit the specific lines captured during inspection.

    answer.html

    PDF · document

    Generated

    Example file from a real run - the skill writes it into your workspace.

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

    About this skill

    Onboarding new developers to a massive codebase usually involves hours of manual walkthroughs or deciphering outdated READMEs. Hypr Repo Guide automates repository discovery by generating structured onboarding guides and targeted Q&A packs directly from source code. It treats code as truth, using a rigorous inspection workflow to capture evidence before making assertions.

    What it does

    • Automated onboarding: Generates a structured guide.md and guide.html following a repository map template.
    • Evidence-backed Q&A: Answers specific technical questions with direct file-and-line citations linked to source passages.
    • Integrity validation: Uses Python helpers to verify that all citations and retained code blocks match the actual source files.
    • Dependency mapping: Scans manifests and documentation to identify setup and test commands without executing them.
    • Source capture: Packages verified evidence into evidence.json and checks.json for auditability.

    How it works

    1. Inventory: Scans the root directory or GitHub URL to map the file structure and read core manifests.
    2. Inspect: Searches identifiers and follows definitions to capture relevant line ranges into an evidence ledger.
    3. Draft: Populates the guide or answer template using strict citations like [path/to/file.py:10-20].
    4. Validate & Render: Runs validate_guide.py to check consistency and render_guide.py to generate the final HTML deliverable.

    Frameworks & tools

    This skill utilizes Python for validation and rendering scripts. It handles local directories and public GitHub repositories, producing Markdown and self-contained HTML files.

    Why this beats prompting it yourself

    Generic prompts often hallucinate file structures or guess how functions work based on their names. This skill enforces a "capture-before-explain" workflow, ensuring every claim is backed by a specific line range in the evidence.json ledger.

    Use cases

    • New hire onboarding: Create a comprehensive repository map for a complex legacy codebase.
    • Feature deep-dives: Generate a focused walkthrough of a specific functional path or module.
    • Technical due diligence: Quickly audit documented setup commands and test coverage across a new repo.
    • Public repo analysis: Inspect public GitHub projects at a specific commit without local cloning.

    Known limitations

    Does not execute project code, run builds, or reveal secrets. Performance is limited by the scope of inspected passages; it cannot confirm the "absence" of features in uninspected files.

    Cannot execute project code or build commands. Requires Python for validation and rendering. Coverage is limited to inspected passages; cannot verify semantic correctness or runtime behavior.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 14 days ago

    • Free with an Agensi account

    Requires

    GitHub
    Python
    Git

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Network Access

    Allowed Hosts

    github
    github.com
    Listed14 days ago

    What's inside

    Frequently Asked Questions

    Unlike standard LLM prompts that guess based on file names, this skill uses a "capture-before-explain" workflow. It extracts actual code snippets into an evidence ledger before writing, ensuring every claim includes a verified file-and-line citation.

    The skill generates a comprehensive repository map (guide.md), a portable HTML version (guide.html) for easy sharing, and an evidence.json file containing all source citations used for the analysis.

    No, the skill is designed for static analysis and discovery; it scans manifests and documentation to identify setup and test commands without ever executing code or running builds on your machine.

    The skill supports local directory analysis and public GitHub repositories. It is optimized for codebases using standard manifest files (like requirements.txt or package.json) and works best with text-based source files.

    Because it generates a verified 'evidence' ledger, you can perform focused deep-dives on specific modules or functional paths, making it ideal for auditing legacy code or performing technical due diligence.

    New in Technical Documentation