Research Toolkit

    1

    Help AI agents write research reports with concrete judgments, traceable sources, and explicit review choices.

    Free

    1 useSecurity scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Research Toolkit

    Example session with this skill installed

    Run a research task on the current state of solid-state battery commercialization for EV manufacturers. I need a brief, source and claim records, and a review step.

    • Read your context and instructions
    • Compiled the research toolkit

    I set up the research task and produced a brief, registered source and claim records, and added a review step whose findings are kept with the task.

    Research Plan: Solid-State Battery Commercialization

    ComponentSpecification
    ObjectiveAssess timeline, energy density claims, and manufacturing barriers for SSBs in EVs.
    ScopeLeading OEMs, technology providers, and 2024-2030 projections.
    Evidence StandardPeer-reviewed journals, manufacturer filings, and pilot plant performance data where available.
    Review StepReview findings are recorded and kept in the task state.
    State Directoryresearch/ssb-report/state/

    Next steps

    1. Collect sources and register findings as source/claim records.
    2. Draft the report from the brief and the registered records.
    3. Run the review step and keep its findings with the task.

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

    About this skill

    Help AI agents write research reports with concrete judgments, traceable sources, and explicit review choices. The current research workflow uses a single entry for new studies, revisions and updates, with the same complete methods for scope, source assessment, claims, analysis, writing and revision. Optional local MCP tools manage task records, stage guidance, registered release-log queries, reviews and delivery checks. Reviews support self/external/independent tiers. Historical lite/full profile values remain compatible; new research does not need a profile choice. MIT-licensed; source and current docs at https://github.com/rrrrrredy/research-toolkit. No general quality advantage over baseline agents has been proven yet; see the repo evaluation status.

    Requires local filesystem for state tracking. Python environment needed for delivery scripts. Does not provide automated source retrieval; depends on agent tools or user-provided access. No general quality advantage over baseline agents has been proven; see the repo evaluation status.

    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

    Requires

    Python
    executable
    Node.js

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Environment Variables
    Listed1 month ago
    Updated21 days ago

    What's inside

    Frequently Asked Questions

    This skill keeps a persistent state and audit trail that records findings and source/claim links, so counter-evidence and unread sources are easier to track than with a plain prompt. It does not guarantee correctness.

    The toolkit is platform-agnostic and designed to work with any OpenAI-compatible HTTP endpoint, as well as developer environments like Codex CLI and Claude Code.

    You will receive the core research framework, a set of structured analysis lenses, and the Python-based verification scripts required to enforce quality gates and audit trails.

    Yes, the toolkit requires a local filesystem for state persistence and a Python environment to execute the delivery verification and source-mapping scripts.

    The current upstream toolkit uses one research entry and the same complete research methods for all tasks. Scope, length and depth follow the request; the agent loads methods, maintains records and completes agreed checks as needed. New requests should omit the profile parameter. Historical lite/full values remain compatible and no longer select separate method sets. Existing task records, review plans and original completion requirements are preserved. See the repository documentation for the current workflow and installation instructions.

    The toolkit uses an 'uncertainty registry' and adversarial reasoning checks to help identify gaps in data, so the final report can note what could not be verified.

    New in Research & Analysis