Asks

1) ClaimCheck API (General Fact Oracle)

  • What it does: Given a natural-language claim + optional URL/context, Cortensor runs the same “fact-check prompt” on multiple miners and returns:

    • true/false/uncertain

    • reasoning + citations

    • dispersion score (how much miners disagree)

  • Users: Any app/site that wants a “verify this claim” button or API.

2) DAO / Governance Proposal Fact Oracle

  • What it does: For a DAO proposal text, miners answer structured questions:

    • “Are the background claims accurate?”

    • “Are key risks omitted?”

  • Validators compare miner answers and produce a governance report (plus PoI-style score).

  • Output: On-chain “fact-check artifact” that DAOs can reference before voting.

3) News & Social Media Post Verifier

  • What it does: Takes a tweet / post URL and asks multiple miners:

    • “Is this claim supported by reputable sources?”

    • “Summarize arguments for/against.”

  • Uses consensus + validator scoring to tag posts as supported, disputed, or unknown.

  • Variant: Browser extension that calls Cortensor in the background.

4) Research Paper / Article Fact Scanner

  • What it does: Upload a PDF or paste an article, then:

    • Split into claims.

    • Run each claim through redundant miners with a “verify & suggest citation” prompt.

  • Output: A report marking:

    • ✅ well-supported

    • ⚠️ ambiguous / needs citation

    • ❌ likely incorrect Great for blog posts, research summaries, or newsletters.

5) Code-Change Risk Oracle

  • What it does: Given a PR diff + repo context:

    • Miners classify risk level, list potential bugs and security issues.

    • Validators compare outputs and score agreement.

  • Use of multi-miner: Same input to many miners with slightly different rubrics (security, performance, readability) → aggregated “risk report” for devs before merge.

6) Contract / Terms-of-Service Explainer + Reality Check

  • What it does: Upload a smart contract or ToS:

    • Miners summarize key points and check for misleading marketing claims vs the actual text.

  • Facts used:

    • “Does the document actually guarantee X?”

    • “Are there hidden fees/clauses?” Great for DeFi users and SaaS buyers.

7) “Agent Watchdog” – AI Output Verifier

  • What it does: Any existing AI agent (trading bot, research agent, etc.) sends its own answer + context to Cortensor.

  • Cortensor miners re-run the task and:

    • confirm / dispute the agent’s answer,

    • assign a trust score,

    • highlight risky recommendations.

  • Variant: On-chain guard that must sign-off (via Cortensor oracle) before a high-risk action is executed.

8) Multi-Source Market Narrative Checker

  • What it does: Input: “Why is BTC down today?” (+ timestamp / tickers).

  • Multiple miners:

    • pull from allowed news/feeds (or provided text),

    • verify whether common narratives (“ETF flows”, “regulation X”) are actually supported.

  • Output: Ranked possible explanations, with a “confidence” derived from miner agreement.

9) “Proof of Reading” / Summary Verifier

  • What it does: For education or compliance:

    • User submits their summary of a document/video.

    • Miners check: “Does this summary faithfully reflect the source?”

  • Use of consensus: Multiple miners rate faithfulness; validator aggregates into a final score and feedback.

10) On-Chain Truth Widgets (Simple Yes/No Oracle)

  • What it does: Minimalist contracts that ask structured, binary questions like:

    • “Did event X occur before date Y?”

  • Cortensor:

    • runs the question through multiple miners,

    • validators score agreement and evidence quality,

    • pushes an attested boolean result on-chain for other protocols to use.

PreviousHackathon #3NextPartnerships

Last updated