config
Point this host at a Linggen — show or set where ling and ling-mem live (loopback, or another machine on your LAN)
Drain the memory review queue — solve each item from evidence, ask the user only when evidence can't settle it
> /plugin marketplace add linggen/linggen-memory > /plugin install linggen@linggen-memory
How it fires
How this command gets triggered: by you, by Claude, or both.
/solveContext preview
What this command does when you run it.
Drain the memory review queue — solve each item from evidence, ask the user only when evidence can't settle it
description: Drain the memory review queue — solve each item from evidence, ask the user only when evidence can't settle it
First call `memory_role`: when it says `role: replica`, reply exactly `Memory is dreamed on the primary (<upstream>)` and stop — review items are solved on the primary.
Otherwise invoke the linggen skill (Skill tool) and treat this message as `/linggen solve` — Solve mode. Follow the skill's Solve runbook: list open items (`memory_issues`, or `ling-mem issues --format json`), then per item try to solve it YOURSELF first — gather evidence at solve time (full rows, git history, code, docs) and write what the evidence settles via `memory_add` + `replace_ids` — asking the user only when evidence can't settle it or a user-voice row is involved (then ONE simple fact question at a time, plain words, with your recommendation; user-voice fixes need `user_directed:true` after the ask), and close each item via `memory_issue_resolve`.
One memory of you, shared by every agent you use — Claude Code, Codex, OpenClaw, and Linggen's own Ling and Yinyue. Tell Claude Code something on Monday; Codex knows it on Tuesday. One local daemon, one store on your machine. No SaaS, no API key, no signup.
Point this host at a Linggen — show or set where ling and ling-mem live (loopback, or another machine on your LAN)
Memory dream pass — judge undreamed days, promote durable signal to long-term, then sweep