Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    hypr research

    1

    Structured research agent that delivers cited Markdown reports, evidence packs, and validated HTML briefs.

    Free

    Security scanned
    hypr research

    hypr research

    Example session with this skill installed

    Research the current state of solid-state battery commercialization for EVs. Use the Standard depth. Save to a new folder named ev-battery-research.

    • Read your context and instructions
    • Compiled the hypr research
    • Generated the document

    I have completed the research on solid-state battery commercialization. The investigation focused on current energy density milestones, pilot production timelines from major manufacturers (Toyota, QuantumScape, Samsung SDI), and remaining technical barriers like dendrite formation and pressure requirements.

    Research Delivery

    FileDescription
    report.mdCited Markdown research brief
    report.htmlPrintable formatted report
    evidence.jsonStructured source metadata and passages
    checks.jsonResults of mechanical validation scripts

    Summary of Findings
    The transition to mass production remains slated for 2027-2028. While lab-scale energy densities exceed 400 Wh/kg [S4, S12], manufacturing scalability and cost parity with liquid electrolytes are the primary bottlenecks. Toyota leads in patent filings, while startups like QuantumScape have moved into B-sample testing with automotive partners [S2, S9].

    Mechanical Validation

    • validate_report.py: PASSED (42 citations verified against 34 sources).
    • render_report.py: PASSED (HTML generated).

    Next steps

    1. Open ev-battery-research/report.html to view the formatted brief and bibliography.
    2. Review evidence.json to inspect the exact passages retained from paywalled technical journals.
    3. Use the

    technical mode if you need a deeper dive into the specific chemical compositions of the solid electrolytes found in S15.

    report.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

    General LLM responses often lack accountability, citations, and structural rigor. When you need a defensible research brief rather than a quick chat, standard prompting fails to track sources or validate consistency.

    This skill implements a formal research workflow that forces the agent to collect evidence, cite specific source IDs, and retain exact passages for audit. It moves beyond simple search by requiring mechanical validation of claims against a structured evidence file.

    What it does

    • Contract-based research: Executes specialized templates for analyst briefs, option comparisons, or technical investigations.
    • Evidence management: Maintains a structured evidence.json file with source IDs, URLs, and retained passages.
    • Budgeting logic: Scales search and page-reading depth across Quick, Standard, or Thorough tiers.
    • Automated validation: Runs Python scripts to check citation consistency and ensure no claims are made without linked evidence.
    • Multi-format delivery: Produces a cited Markdown report and a printable HTML document.

    How it works

    1. Define the contract: The agent identifies the research type (analyst, comparison, or technical) and sets a depth budget based on your requirements.
    2. Collect evidence: It searches and reads pages, assigning stable IDs (S1, S2) and recording whether a source was fully read or just a snippet.
    3. Draft and cite: The agent writes report.md using the selected template, placing citations like [S1] directly next to supported claims.
    4. Validate and render: It executes bundled Python scripts to perform mechanical consistency checks and generate the final HTML report.

    Frameworks & tools

    The skill requires Python 3.10+ for the validation and rendering scripts. It relies on the host agent's internal web search and page-reading tools to populate the evidence pack.

    Why this beats prompting it yourself

    Manual prompting often results in "hallucinated" citations or vague references to general knowledge. This skill enforces a structured data contract between the research process and the final output, using external scripts to verify that every citation actually exists in the collected evidence.

    Use cases

    • Market analysis: Compare three competing SaaS products with cited pricing and feature sets.
    • Technical due diligence: Investigate a specific library's architecture and known security vulnerabilities.
    • Evidence reviews: Summarize a collection of provided PDFs with exact passage retention for auditing.
    • Decision support: Evaluate geographical locations for a new office based on specific economic constraints.

    Known limitations

    Requires an agent environment capable of executing Python. It cannot verify the underlying truth or authenticity of a source, only the mechanical consistency of citations.

    Requires Python 3.10+ and agent tool access for search/reading. Mechanical validation checks for citation consistency but does not verify the factual accuracy or authenticity of sources.

    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

    Python

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Network Access
    Listed14 days ago

    What's inside

    Frequently Asked Questions

    Standard AI responses often lack verifiable proof; this skill enforces a formal 'evidence contract' by recording specific source IDs and passages in a JSON file to ensure every claim is auditable.

    The skill includes a structured research framework, Python-based validation scripts for citation checking, report templates (Analyst, Comparison, Technical), and an automated HTML renderer.

    This skill requires an agent environment with Python 3.10+ support and access to web-searching/page-reading tools to populate its evidence database.

    Yes, you can choose between Quick, Standard, or Thorough tiers to control the depth of page-reading and search iterations based on your time and token constraints.

    The skill delivers a structured 'evidence.json' pack, a cited Markdown report, and a professionally formatted HTML brief ready for printing or presentation.

    The skill uses mechanical validation to ensure every citation matches a real source in the evidence file, effectively preventing the 'hallucinated' citations common in standard LLM outputs.

    New in Research & Analysis