External Brain System (EBS)

The description states that External Brain System (EBS) is a system designed to preserve AI conversation state across models and providers by creating persistent, recoverable working memory for complex projects. It claims to enable durable collaboration with AI assistants through structured records that distinguish between confirmed decisions, rejected alternatives, rules, and other knowledge types.

What changed: The author describes EBS as evolving from an experimental Gmail-based system into a portable Python package with command-line interface, using GPT-5.6 and Codex for development.

The single most important open question: Is there evidence of actual usage or adoption beyond the author's own development work? The description does not indicate any customers, users, or external validation of EBS functionality.

The description states that EBS is a system that "moves the durable working state outside the individual conversation" and captures project knowledge as structured records with stable identities, versions, relationships, provenance, lifecycle states, authority roles, recovery information, current focus, exact checkpoints, and next actions.

EBS distinguishes between different kinds of information including confirmed decisions, decision candidates, rejected alternatives, rules and protected constraints, reusable knowledge, processes and agents, mutable working copies, immutable snapshots, execution evidence, maintenance records, append-only logs, open tasks, and more.

The system creates two complementary layers: a narrative layer preserving reasoning path (project objective, important findings, active decisions, decision rationale, rejected alternatives, unresolved questions, current focus, exact checkpoint, immediate next action) and a complete current record state required for safe continuation.

EBS is described as having both a Gmail-based implementation that served as an early architectural test and a portable repository version built with Python, Markdown, YAML, JSON schemas, SQLite, and FTS5 search.

The description states EBS positions itself as a solution to the fundamental limitation of AI conversations being temporary. It claims to turn finite AI conversations into persistent, recoverable states that preserve decisions, reasoning, tasks, rules, provenance and key checkpoints across models and providers.

The author's claim evolution shows a progression from recognizing the problem (temporary conversations losing important context) to developing a solution (EBS) that works in practice before being extracted into portable code. The system is positioned as enabling "functionally unbounded continuity through persistent external state, selective activation, governed reconstruction, and controlled continuation."

The author states EBS does not claim to create infinite physical context window but instead provides functional continuity through persistent state management.

Not evidenced. The description does not identify specific target customers or ideal customer profiles beyond the author's own use case of working with AI for software, business, research, planning, writing and creative projects.

Not evidenced. The description does not contain any information about pricing, revenue models, monetization strategies or business model details.

The description states EBS was built using GPT-5.6 and Codex with Python, Markdown, YAML, JSON schemas, SQLite, FTS5 search, git, github, gmail, actions, chatgpt, codex, and other technologies. It includes a portable repository version with command-line interface, filesystem and SQLite storage adapters, record and relationship validation, transition-package generation, isolated target-chat evaluation, ChatGPT export normalization, semantic candidate extraction, reconciliation boundaries, automated tests, CI and jury workflows, and executable documentation.

The author describes using Gmail as the initial durable primitive mapping EBS concepts to existing infrastructure before building a custom technical platform. The system supports both mutable working copies (drafts) and immutable snapshots (messages), with labels serving as project and record views.

Not evidenced. The description does not contain any information about revenue, customers, user adoption, market traction or maturity indicators beyond the author's own development work.

Not evidenced. The description does not identify competitors, market positioning relative to other tools, or competitive landscape information.

The description states that storing text is easy but reliable continuity is hard. It identifies challenges including deciding what "memory" actually means, transition fidelity, privacy concerns, and ensuring trustworthy target-model evaluation without access to private environments.

Key risks include:

• The system appears to be primarily self-developed with no external validation or user feedback
• No evidence of customers, users or market traction
• The demonstration is based on a sanitized 37-record version excluding private information
• The system's functionality has not been independently verified
• The author's own account may overstate the practical utility beyond their personal use case

1. What specific problems have you encountered in practice with AI conversation state management that EBS solves?
2. How does EBS handle conflicts between different versions of information or decisions?
3. Can you demonstrate actual usage patterns beyond your own development work?
4. What are the practical limitations of the current implementation for real-world use cases?
5. How do you plan to scale beyond a single developer's use case?
6. What validation methods have you used to ensure transition fidelity and recovery accuracy?
7. How does EBS handle privacy and data protection in actual usage scenarios?
8. What are your plans for expanding beyond the current technical implementation?