Works with the AI tools you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+22 more

    verified handoff

    1

    Ensure seamless session transitions by verifying git state, test results, and uncommitted changes across AI agents.

    $5

    Secure checkout via Stripe

    30-day refund guarantee

    Converts to your local currency at checkout

    Security scanned
    verified handoff

    verified handoff

    Example session with this skill installed

    I'm out of tokens on this session. Save a handoff for the feature-auth branch and run the npm test command to verify the login flow work.

    • Read your context and instructions
    • Compiled the verified handoff

    I have captured the current session state and verified the evidence. A handoff file has been created with 4 verified claims and 1 unverified belief regarding the session state.

    Handoff · saved ✅

    .handoff/session_auth_refactor.md · branch feature-auth @ a1b2c3d

    ClaimsnStatus
    Verified4Commits, File Hashes, Tests
    Believed1Logic assumptions
    Failed0-

    Next step: Implement the JWT refresh token logic in src/auth/service.ts.

    Handoff Brief

    The session focused on the core authentication middleware. The following state is recorded:

    Done:

    • [verified: commit a1b2c3d] Initial middleware structure.
    • [verified: test "npm test" passed] Login controller unit tests.
    • [verified: file src/middleware/auth.ts matches hash] Core logic implemented.

    In Progress:

    • Uncommitted changes in src/auth/service.ts regarding token expiration.
    • Decision: Used a custom header instead of Cookies to simplify the initial mobile integration.

    Pointers:

    • See docs/auth-flow.md for the sequence diagram.
    • Run npm run dev to test the mock login endpoint.

    Next steps

    1. Switch to your new agent session.
    2. Run node scripts/handoff.mjs check --repo . --file .handoff/session_auth_refactor.md --run-tests.
    3. Follow the recommended plan to complete the refresh token logic.

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

    About this skill

    Development sessions often end abruptly due to token limits, context window exhaustion, or the need to switch between different AI tools. When you resume, the new agent usually guesses the state of the repo, often building on stale assumptions or hallucinated progress. This skill provides a verifiable bridge between sessions, ensuring no context is lost and every claim is backed by the current state of the repository.

    What it does

    • Evidence collection scans git commits, file hashes, and test results to prove what was actually accomplished.
    • Context checkpointing records uncommitted work and the specific reasoning behind architectural decisions that code alone cannot explain.
    • Truth verification re-validates handoff claims upon resumption, alerting you if files changed or tests started failing in the interim.
    • Cross-agent migration generates standardized Markdown and JSON reports that Claude Code, Cursor, or Gemini CLI can ingest to resume work instantly.

    How it works

    1. Save session by running node scripts/handoff.mjs save to capture current progress, test status, and uncommitted changes.
    2. Review claims as the skill generates a report tagging items as verified, believed, or unverified based on hard evidence.
    3. Resume work by running node scripts/handoff.mjs check in a new session to verify which previous claims still hold true against the live repo.
    4. Follow the brief to see a summary of confirmed progress and the immediate next step in the development plan.

    Frameworks & tools

    This tool requires Node.js 18+ and Git. It operates locally within any git repository and is compatible with terminal-based AI agents like Claude Code, as well as IDE-integrated tools like Cursor.

    Why this beats prompting it yourself

    Manual handoffs are prone to hallucinations where the agent claims a feature is finished when the tests actually fail. This skill automates the verification process, using file hashes and shell execution to ensure the "Next Step" is based on reality rather than session memory.

    Use cases

    • Resuming a complex refactor after hitting daily LLM usage limits.
    • Transferring work from a CLI agent to a GUI-based IDE agent.
    • Checking out a coworker's branch and instantly seeing their "verified" progress.
    • Documenting architectural "why" decisions that are not captured in commit messages.

    Known limitations

    Requires Node.js 18+ and a local git environment. Verification of claims requires the skill script to be accessible to the agent in the current environment.

    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

    30-day refund

    Not a fit? Get your money back

    Trust & safety

    Security scanned

    Verified clean 3 days ago

    • Passed all security checks, Safe to install

    Listed3 days ago

    What's inside

    Frequently Asked Questions