菜单
Load the Nowledge Mem Working Memory briefing at Proma session start, after a break, or when recent context would help the current task.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/nowledge-co/community --skill read-working-memory命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Read the user's Working Memory briefing when current priorities, recent activity, or active focus areas matter.
Read your daily Working Memory briefing to understand current context. Contains active focus areas, priorities, unresolved flags, and recent knowledge changes. In Grok Build, invoke this on the first relevant turn because passive startup-hook output is not model context.
Check Nowledge Mem setup, detect your agent, and guide native connector setup. Use when the user asks about setup, configuration, or when memory tools aren't working as expected.
基于 SOC 职业分类
正在显示 SKILL.md
| name | read-working-memory |
| description | Load the Nowledge Mem Working Memory briefing at Proma session start, after a break, or when recent context would help the current task. |
Read Context Bundle for full Proma startup context when available: owner identity, AI Identity, active scope, active rules, and Working Memory. Read Working Memory alone when you only need current priorities.
Proma's lifecycle hook also writes startup context into the workspace CLAUDE.md inside a nowledge-mem:start / nowledge-mem:end block. Treat that block as the automatic startup layer; call MCP or CLI when you need fresh or more specific retrieval during the session.
Session start: beginning a new Proma conversation, returning after a break, or when recent context would help.
During session: user asks about current work, references recent decisions, or says "what was I working on?"
Skip if: already loaded this session, user wants a fresh start, or the task is isolated and trivial.
Primary (MCP, full startup context):
mcp__nowledge-mem__read_context_bundle
Lightweight MCP:
mcp__nowledge-mem__read_working_memory
CLI Context Bundle fallback:
nmem --json context --source-app proma
Working Memory fallback:
nmem wm read
# or with JSON output:
nmem --json wm read
Add --space "<space name>" for space-aware lanes. Multi-agent orchestrators can set NMEM_AGENT_ID="<agent-slug>" before launching Proma. Add NMEM_SPACE only when that whole run should override the identity's default space. Use NMEM_HOST_AGENT_ID only for advanced external aliases.
The briefing includes: active focus areas ranked by activity, flagged priorities, unresolved flags, recent knowledge-base changes, and deep links to specific memories. Use nowledgemem:// links to reference specific memories when relevant.
If MCP tools are unavailable, verify ~/.proma/agent-workspaces/default/mcp.json has the nowledge-mem server configured with key "servers". For CLI fallback, install via pip install nmem-cli or use the Nowledge Mem desktop app.