Cycling through all six. Tap any point to stop.

Measured on six things

AI memory layer that connects your tools and remembers yo...

Growth

Extremely early beta with waitlist only; tiny social mentions and no reviews or community yet.

Screenshots

1 / 6

innernet

About this app

Innernet is a persistent, versioned memory layer for your digital life. It quietly observes your work across AI tools like ChatGPT, Claude, Cursor, and Perplexity, capturing decisions, preferences, and context so that every tool you use stays in sync. Think of it as a searchable, evolving 'second brain' that lives at the center of your workflow. With row-level security, hashed tokens, OAuth 2.1, and optional field-level encryption, your memory stays private and yours. It offers MCP, REST API, and a Git-like anchor for code repos, making it ideal for developers, creators, and researchers who want to stop re-explaining context and just build.

Key features

1MCP + REST API for live context sync across ChatGPT, Claude, Cursor, Perplexity and more.
2Versioned, branchable memory maps with commit logs and capture protocols per project.
3Row-level security, hashed tokens, OAuth 2.1, optional field encryption, GDPR controls.

Target customer profile

Drawn from the product itself, not from a survey.

Demographic

Solo developers and indie hackers

Pain points

Losing context between coding sessions; repeating setup instructions to AI tools

Primary needs

Seamless memory persistence across development tools; minimal setup

Demographic

Content creators and writers

Pain points

Scattered notes and drafts; losing voice and research threads

Primary needs

Centralized, versioned memory for ideas and brand voice; cross-tool sync

Demographic

Researchers and analysts

Pain points

Disconnected sources and synthesis; difficulty tracking evolving findings

Primary needs

Living document that auto-structures research; memory that grows with exploration

Scout Review

Written by AI from measured evidence, scored out of 100.

Innernet offers a compelling vision for a versioned, privacy-first memory layer that syncs context across AI coding and research tools. Its strength lies in thoughtful security details and seamless MCP/REST integrations that directly address the pain of losing context between sessions. The polished site and quick-setup promise make it attractive for solo developers and creators. However, being in early beta with no pricing, demos, or user feedback yet means it remains unproven. It sits ahead of basic note-taking alternatives in AI-tool depth but behind established players in maturity and ecosystem.

Measure by measure

Usability

Fast MCP setup and clear docs make onboarding sound painless for devs already using supported tools.

Design

Polished, modern landing with elegant code examples and memory visualizations that feel thoughtful and on-brand.

Speed

Site feels snappy with lightweight presentation; real-time sync claims untested publicly.

Security

Strong-sounding claims around row-level security, hashed tokens, and OAuth match the app description exactly.

Accessibility

Responsive design visible but zero accessibility documentation or standards mentioned anywhere.

Growth

Extremely early beta with waitlist only; tiny social mentions and no reviews or community yet.

Improvement areas

Add public pricing page and live demo to reduce friction for early adopters.
Build community presence (GitHub, Discord) and gather initial user feedback.
Address accessibility gaps and pursue basic compliance documentation.

Design & Usability

Innernet's landing page nails a clean, developer-friendly aesthetic with well-crafted code snippets and memory timeline visuals that communicate the 'persistent second brain' concept effectively. The design feels modern and consistent, avoiding clutter while highlighting integrations with tools like Claude and Cursor. Usability shines in the promised frictionless MCP connection — sign in once via OAuth and context flows across sessions without extra installs. Documentation at /docs provides clear starting points for developers. However, the absence of a live demo or public pricing page keeps it from feeling fully ready for broad adoption. Overall, it positions well for its target solo devs and researchers who hate re-explaining context.

Speed & Security

The site itself performs smoothly with no noticeable load issues, supporting quick browsing of features. Real-time memory sync via MCP and REST API is central to the pitch, though no independent benchmarks exist yet. Security stands out with concrete, verifiable claims: row-level database enforcement, hashed access tokens, OAuth 2.1, optional field encryption, and GDPR-compliant export/delete controls. These align directly with the product description and privacy section. Still early days mean no third-party audits or SOC 2 reports, which is typical but leaves some uncertainty. For a memory layer handling personal context across AI tools, the transparency on data handling is a solid foundation compared to many indie tools.

Accessibility & Growth

Accessibility receives minimal attention — the responsive layout and readable text are baseline positives, but there's no mention of WCAG compliance, alt text practices, or screen reader support. This is common for very new products but represents a gap versus more mature peers. Growth is the weakest area: as of July 2026 the product is in pre-launch beta with a waitlist, planned launch around early July, and only scattered social mentions on LinkedIn and Reddit. No GitHub activity, review platform presence, or community metrics exist for this specific innernet (distinct from the unrelated 2021 WireGuard project). Trajectory depends entirely on successful beta execution and word-of-mouth among indie hackers.

Conclusion

If you're tired of repeating yourself to every new AI chat, innernet is worth joining the waitlist to test once it opens. Just temper expectations — this is very much a product still finding its footing in a crowded second-brain space.

Suggestions for improvement

Publish pricing and a live demo to convert waitlist interest faster.
Add a public changelog and GitHub presence for developer trust.
Implement basic accessibility audits and document WCAG efforts.
Secure an independent security review or SOC 2 path as usage grows.

How we score

Alternatives comparison

Named competitors, point by point. Nobody paid to appear here or to be left out.

	innernet	Mem.ai	Notion AI	Rewind.ai
Live AI tool integration (MCP/REST sync across multiple tools)	Yes - MCP endpoint + REST for ChatGPT, Claude, Cursor, Perplexity etc.	Limited - AI note-taking but no deep multi-tool live sync	Partial - Notion AI workspace but not automatic live memory from external tools	No - Screen/meeting recording, local-only, no developer tool focus
Versioned/branchable memory with commit history	Yes - Git-like branches, merges, commit logs on context maps	Basic versioning on notes	Page history but not structured branching for memory	Timeline recording but no versioned branching
Privacy & security features (row-level, encryption, OAuth)	Strong claims: row-level security, hashed tokens, OAuth 2.1, opt-in encryption	Standard encryption and privacy controls	Enterprise SSO, encryption, SOC 2	Local-only storage emphasis