M-T-D-N/agentmemory-codex-windows
by Various
AgentMemory for Codex on Windows: persistent project-scoped memory, keyword/graph retrieval, canonical filters, provenance-preserving graph tools, four managed hooks, optional loca
MCP
M-T-D-N/agentmemory-codex-windows
Added 9 Oct 2026
Overview
AgentMemory for Codex on Windows provides persistent, project-scoped memory for Codex agents. It supports keyword and graph retrieval, canonical filters, provenance-preserving graph tools, four managed hooks, optional local Qwen graph extraction, and recoverable managed updates. It is an unsigned technical preview with per-release validation results and latency limits.
Best for
Best for
Developers using Codex on Windows who need persistent, project-scoped memory with graph retrieval.
Use cases
- Persist project-specific context across Codex sessions
- Retrieve relevant memory via keyword or graph search
- Extract knowledge graphs locally with optional Qwen
How to use
Tested with
Cursor, Continue, ChatGPT
Notes
AgentMemory for Codex on Windows provides persistent, project-scoped memory for Codex agents. It supports keyword and graph retrieval, canonical filters, provenance-preserving graph tools, four managed hooks, optional local Qwen graph extraction, and recoverable managed updates. It is an unsigned technical preview with per-release validation results and latency limits.
2 stars on GitHub. Last updated 2026-10-08. Licensed Apache-2.0.
Use cases
- Persist project-specific context across Codex sessions
- Retrieve relevant memory via keyword or graph search
- Extract knowledge graphs locally with optional Qwen
Pros
- Persistent project-scoped memory keeps context isolated per project
- Provenance-preserving graph tools maintain traceability
- Recoverable managed updates reduce risk during upgrades
Cons
- Unsigned technical preview, not production-ready
- Limited to Windows and Codex environment
- Latency limits documented per release may constrain large graphs
Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.
Pros
- Persistent project-scoped memory keeps context isolated per project
- Provenance-preserving graph tools maintain traceability
- Recoverable managed updates reduce risk during upgrades
Cons
- Unsigned technical preview, not production-ready
- Limited to Windows and Codex environment
- Latency limits documented per release may constrain large graphs
Get the free Developer’s Field Guide
A 27-page field guide to the AI coding workflow with Claude. Claude Code, MCP servers, the prompt patterns that work, and what to delegate. Free.
Enter your work email. We send it straight over, plus a few short notes worth knowing. Unsubscribe any time.
