GoodMemory

The description states that GoodMemory is a memory system designed to give AI products and coding agents:

• The ability to remember selected facts, preferences, experiences, and project knowledge
• Recall relevant context for current tasks
• Assemble evidence-backed context packs for next model turns
• Inspect why memories were selected
• Accept feedback, revise stale knowledge, forget individual memories, export data, or delete everything

It installs into Codex and Claude Code using hooks, lifecycle management, local SQLite storage, read-only MCP inspection, and opt-in writeback. The system supports TypeScript, HTTP, Python, AI SDK, OpenAI Agents SDK, or any MCP-compatible client.

The core is written in TypeScript and runs on Bun, with SQLite as the default local storage engine. PostgreSQL and custom adapters are also supported for larger deployments.

It includes a public API with functions such as:

• remember
• recall
• buildContext
• feedback
• forget
• exportMemory
• deleteAllMemory

The system also offers an optional Inspector that provides categorized memory, candidate review, recall traces, revision/deletion controls, and audit events.

Inference This is a software layer intended to be embedded into AI agents or applications to manage their memory lifecycle explicitly and locally.

The description states:

• GoodMemory started from the principle that durable memory should be a product layer of its own
• It positions itself as local-first, explicit, auditable, and reversible
• It contrasts with current practices like resending entire histories, bolting on vector databases, or handing user memory to opaque hosted services

The project claims:

• It gives AI apps and coding agents a “complete memory loop”
• It supports local-first installation without requiring accounts or hosted services
• It allows for governance of memory through hooks, MCP, and explicit module boundaries
• It integrates with Codex and Claude Code via setup flows and lifecycle hooks

Inference The positioning is that of a memory infrastructure tool, not an end-user product. It targets developers building AI agents or applications who want to manage memory in a more controlled way.

The description states:

• GoodMemory is intended for application developers integrating into AI products
• It supports integration through TypeScript, HTTP, Python, AI SDK, OpenAI Agents SDK, or MCP-compatible clients
• It works with Codex and Claude Code, suggesting a focus on AI coding tools
• It supports local-first deployment, implying use cases where data sovereignty or performance matters

Inference The primary ICP appears to be developer tooling teams building AI agents or applications that require memory management. Secondary users may include developers working in environments where local storage and auditability are important.

The description states:

• GoodMemory is published as an open-source npm package (v0.6.0)
• It supports local-first installation with no hosted service required
• Integration paths include TypeScript, HTTP, Python, AI SDK, MCP

There is no mention of:

• Pricing models
• Subscription or licensing fees
• Revenue streams
• Paid features or tiers

Inference The business model is not evident. It may be open-source with a potential freemium or enterprise model, but this is not stated.

The description states:

• Built with bun, codex, mcp, node.js, openai, postgresql, python, react, sqlite, typescript, vite
• Core written in TypeScript and runs on Bun
• Uses SQLite as default, with support for PostgreSQL and custom storage adapters
• Retrieval path combines:
  • Generalized BM25
  • Entity signals
  • Reciprocal-rank fusion
  • Temporal and relation-aware planning
  • Optional embeddings and provider reranking
• Public API includes functions like remember, recall, buildContext, etc.
• Supports CLI setup flows, MCP, HTTP, Python, and TypeScript integrations

Inference The technical stack is modern and developer-focused. The system supports multiple integration methods and has a modular architecture, suggesting scalability and extensibility.

The description states:

• Published goodmemory@0.6.0 as an open-source npm package
• Supports multiple Node versions, storage backends, browser flows, and thousands of tests
• Includes release verification covering multiple environments
• Has opt-in benchmark declarations (LoCoMo, BEAM 100K, MemoryAgentBench)
• Is submitted to the OpenAI 2026 hackathon

There is no mention of:

• Customers or users
• Revenue or monetization
• Product adoption or usage metrics
• Product-market fit

Inference This is a prototype or early-stage product, likely built for a hackathon. There is no evidence of traction, revenue, or customer base.

The description does not mention:

• Direct competitors
• Market positioning relative to existing memory systems or AI agent frameworks
• Comparison with tools like vector databases, memory management libraries, or agent frameworks

Inference No competitive context is provided. It’s unclear whether this product addresses a gap in the market or competes with existing solutions.

The description states:

• The hardest challenge was deciding what deserves to become durable memory
• It separates production retrieval from benchmark experiments
• It keeps answer models and judges independent
• It adds strict public-claim gates to prevent internal diagnostics from becoming marketing claims

Inference This is a pre-product concept, likely built for demonstration purposes. It lacks commercial viability or traction indicators.

1. What is the actual use case or problem you are solving? Is there a real-world application of this system?
2. How do you plan to monetize this product if it's open-source?
3. Are there any early adopters or pilot users who have tested this in production?
4. What are the performance and scalability limitations of the current implementation?
5. How does GoodMemory handle data privacy, especially with local-first storage?
6. What is your roadmap for moving from prototype to a commercial product?