Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    crypto workbench

    1

    A staged workflow for implementing and reviewing cryptography from papers, standards, and RFCs.

    Free

    1 useSecurity scanned
    crypto workbench

    crypto workbench

    Example session with this skill installed

    I need to implement the OPAQUE asymmetric PAKE protocol based on draft-irtf-cfrg-opaque-11. Start with the crypto-init stage.

    • Read your context and instructions
    • Compiled the crypto workbench

    I have initialized the .cryptoworkbench/ directory and created the project structure. I've recorded the target as the OPAQUE protocol (draft-irtf-cfrg-opaque-11).

    Crypto Workbench Initialization

    ItemStatusDetails
    Project RootCreated.cryptoworkbench/
    Target SpecIdentifieddraft-irtf-cfrg-opaque-11
    Open QuestionsCreatedopen_questions.md initialized with draft versioning
    Decision LogCreateddecisions.md ready for library choices

    I have prepared the workbench. Before we move to crypto-intake, we need to address the initial parameters for this implementation.

    Open Questions:
    1.

    Implementation Style: Should this be a research prototype, a reference-style implementation, or a library-backed production-oriented version?
    2. Target Language: Do you have a preferred language (e.g., Python, C++, Rust) for the implementation and testing?
    3.

    Group Selection: OPAQUE requires a prime-order group. Should we default to Ristretto255 or a specific NIST curve?

    Next steps

    1. Answer the open questions above to update decisions.md.
    2. Run crypto-intake to fetch and scan the RFC draft text.
    3. Begin the crypto-paper-review to identify the required primitives (HKDF, Hash-to-Group, etc.).

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

    About this skill

    Moving from a whitepaper or RFC to a production-ready cryptographic implementation is prone to subtle errors and security flaws. This skill enforces a rigorous, staged workflow to ensure every primitive, notation, and security assumption is documented before a single line of code is written.

    What it does

    • Source extraction scans PDFs and RFCs to build notation maps and identify ambiguities in cryptographic specifications.
    • Library mapping matches required primitives to established libraries like OpenSSL, NaCl, or BoringSSL to avoid "rolling your own."
    • Spec construction generates detailed implementation specs, ambiguity logs, and test plans based on peer-reviewed research.
    • Staged implementation guides the transition from plan to code, ensuring tests are written against the spec's edge cases.
    • Security review performs a structured reading of the implementation to identify logic errors and deviations from the target standard.

    How it works

    1. Initialize the project using crypto-init to create the local workbench directory and define the target algorithm or paper.
    2. Ingest the source material via crypto-intake and crypto-paper-review to extract parameters and threat models.
    3. Map and Plan by running crypto-library-map and crypto-plan-review to select tools and approve the architecture.
    4. Execute the build through crypto-implement and crypto-security-review to produce the final, reviewed code.

    Frameworks & tools

    This skill requires Python 3 with PyYAML and pypdf. It references a local registry of algorithms and math families to ensure standard compliance.

    Why this beats prompting it yourself

    Generic prompts often ignore the subtle edge cases in academic papers or suggest unsafe hand-rolled crypto. This skill enforces a mandatory review gate between specification and coding, preventing the LLM from making silent assumptions about encodings or security parameters.

    Use cases

    • Paper implementation translating a new research paper into a verified Python or C prototype.
    • Standard compliance auditing an existing codebase against a specific RFC or NIST standard.
    • Library migration planning the move from legacy primitives to modern, library-backed implementations.

    Known limitations

    The security review is a static reading of code and does not replace formal verification or a professional audit. Online access for fetching papers requires explicit per-request user consent.

    Security reviews are manual code readings, not formal proofs. Requires Python 3 and PyYAML. User must provide explicit consent for each external URL fetch.

    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

    1 use

    Times used to date

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 1 month ago

    • Free with an Agensi account

    Permissions

    Read Files
    Write Files
    Terminal / Shell
    Listed1 month ago

    What's inside

    Frequently Asked Questions

    It solves the 'translation gap' by enforcing a rigorous, multi-stage workflow that maps academic notations and RFC specs to code, preventing common implementation errors and unsafe 'hand-rolled' crypto primitives.

    The skill is designed for Python-based AI agents and requires Python 3, PyYAML, and pypdf; it functions as a structured workflow orchestrator that guides your agent through specific review gates.

    You receive the full suite of workbench commands (like crypto-init and crypto-intake), the local algorithm registry, and the structured workflow templates for source extraction and security review.

    While it identifies logic errors and deviations from standards, it is a static review tool and does not provide mathematical formal verification or replace a professional third-party security audit.

    Setup is straightforward: once added to your agent, you run the `crypto-init` command to build your local workbench directory and link your source PDF or RFC.

    New in APIs & Backend