// Recall and store long-term memory in Codex using the memory-lancedb-pro MCP tools. Use when the user asks what happened before, wants to continue prior work, asks to remember a fix or preference, or types `@memory`.
| name | memory |
| description | Recall and store long-term memory in Codex using the memory-lancedb-pro MCP tools. Use when the user asks what happened before, wants to continue prior work, asks to remember a fix or preference, or types `@memory`. |
| metadata | {"short-description":"Recall and store long-term memory for this repo"} |
Use this skill to make persistent memory feel like a first-class Codex workflow.
$memory explicitly invokes this skill./memory should activate this skill in a new thread after reload.@memory or @记忆 as a user-facing alias for this skill even though it is not a native app-backed @ chip.memory_recall if prior fixes, decisions, or preferences may matter.project:new-project unless there is a better repo-specific slug in the current task.memory_extract_and_store with a short summary transcript.memory_store only for a single short atomic memory.memory_health if recall or writes fail.@memory recall <query>: call memory_recall@memory store <text>: call memory_store@memory extract <summary>: call memory_extract_and_store@memory list: call memory_list@memory stats: call memory_stats@memory health: call memory_health@memory forget <id>: call memory_forget