Skip to content

TERMS_OF_USE.md

Latest commit

 

History

History
77 lines (54 loc) · 2.72 KB

TERMS_OF_USE.md

File metadata and controls

77 lines (54 loc) · 2.72 KB
Raw

InfoMesh — Terms of Use

Last updated: February 2026

1. Acceptance

By using InfoMesh software ("the Software"), you agree to these terms.

2. What InfoMesh Is

InfoMesh is a decentralized, peer-to-peer search engine designed for LLMs. Each node independently crawls, indexes, and searches the web. There is no central server — your node is both client and server.

3. Content and Crawling

3.1 Crawled Content

  • InfoMesh caches web content temporarily for search indexing purposes.
  • All crawling respects robots.txt directives and site terms of service.
  • Cached content is stored locally with a default TTL of 7 days.
  • Search results return snippets only — not full page content.

3.2 Content Responsibility

  • You are responsible for content your node crawls and serves.
  • Do not configure your node to intentionally crawl illegal content.
  • InfoMesh includes a domain blocklist; do not disable it.

3.3 Takedown Compliance

  • DMCA takedown requests propagated via the DHT must be honored within 24 hours.
  • GDPR deletion records in the DHT must be honored immediately.
  • Failure to comply may result in trust score reduction and network isolation.

4. Network Participation

4.1 Fairness

  • The credit system rewards contribution and penalizes free-riding.
  • Nodes with negative credit balance may be rate-limited.
  • Intentional credit farming (e.g., fabricated crawl results) is prohibited.

4.2 Trust

  • Node trust is computed from four signals: uptime, contribution, audit pass rate, and summary quality.
  • Nodes failing 3 consecutive random audits are isolated from the network.
  • Tampered content (modified hashes) violates network integrity.

4.3 Resource Usage

  • Configure resource profiles to match your hardware and network capacity.
  • Do not exceed reasonable crawl rates for the domains you visit.
  • Default limits (60 URLs/hr, 5 Mbps upload) exist for a reason.

5. Privacy

  • No central telemetry. InfoMesh never phones home.
  • Search queries stay local or are routed via the DHT with no central logging.
  • Your node's crawl history is stored locally in ~/.infomesh/.
  • P2P traffic is encrypted via libp2p noise protocol.

6. LLM Summarization

  • AI-generated summaries are labeled as such.
  • Every summary includes a content_hash linking to the original source.
  • Original URLs are always provided alongside AI summaries.
  • Summaries are not a substitute for the original content.

7. No Warranty

InfoMesh is provided "as is" without warranty of any kind. See the MIT License for full terms.

8. Changes

These terms may be updated. Significant changes will be announced via the project repository.

Questions? Open an issue at https://github.com/dotnetpower/infomesh