← 返回
未分类 Key 中文

Agentnotes

Log OpenClaw tasks to AgentNotes for SparkNotes rollups (success, failures, what happened). Use after cron jobs, channel replies, or multi-step sessions. Req...
将OpenClaw任务记录到AgentNotes,以便SparkNotes汇总(成功、失败、发生的情况)。适用于cron任务、频道回复或多步骤会话后。需求...
mattmerrick
未分类 clawhub v0.1.0 1 版本 99248.1 Key: 需要
★ 0
Stars
📥 132
下载
💾 0
安装
1
版本
#latest

概述

AgentNotes for OpenClaw

Send high-level job summaries to AgentNotes. Raw logs are rolled up hourly/daily into SparkNotes; you do not need to store every line forever.

Setup (once)

  1. AgentNotes dashboard → create agent (note the slug) → copy env vars from Connect.
  2. Install this skill (see dashboard Connect → OpenClaw tab, or run install-skill.ps1 / install-skill.sh from the repo).
  3. Add env to ~/.openclaw/openclaw.json:
{
  skills: {
    entries: {
      agentnotes: {
        enabled: true,
        env: {
          AGENTNOTES_API_KEY: "an_...",
          AGENTNOTES_AGENT_ID: "your-slug",
          AGENTNOTES_BASE_URL: "https://your-app.vercel.app",
        },
      },
    },
  },
}
  1. Restart OpenClaw / new session. Verify:
node {baseDir}/scripts/verify.mjs

Recommended: one command per task

Always pass a plain-English summary (this becomes the CEO-facing SparkNote):

node {baseDir}/scripts/task.mjs --summary "Replied to 2 Telegram chats, scheduled 1 meeting"

With an extra step log:

node {baseDir}/scripts/task.mjs --summary "Inbox sweep done" --message "Processed 12 threads" --step inbox

On failure:

node {baseDir}/scripts/task.mjs --summary "Send failed" --failed --error "SMTP timeout"

Manual multi-step flow

export AGENTNOTES_RUN_ID=$(node {baseDir}/scripts/start-run.mjs)
node {baseDir}/scripts/log.mjs "Tool: calendar" --step tools
node {baseDir}/scripts/complete-run.mjs --summary "Updated 3 events"

When to log

  • After handling a user message on any channel
  • After a cron / heartbeat completes
  • After a skill or tool chain finishes (success or failure)

Dashboard

$AGENTNOTES_BASE_URL/agents — SparkNotes tab is the long-term record.

版本历史

共 1 个版本

  • v0.1.0 当前
    2026-06-04 13:55

安全检测

腾讯云安全 (Keen)

队列中

腾讯云安全 (Sanbu)

队列中

🔗 相关推荐

ai-intelligence

self-improving agent

pskoett
捕获经验教训、错误和纠正,以实现持续改进。使用时机:(1)命令或操作意外失败;(2)用户纠正……
★ 4,061 📥 799,434
security-compliance

Skill Vetter

spclaudehome
AI智能体技能安全预审工具。安装ClawdHub、GitHub等来源技能前,检查风险信号、权限范围及可疑模式。
★ 1,219 📥 266,796
ai-intelligence

Self-Improving + Proactive Agent

ivangdavila
自我反思+自我批评+自我学习+自组织记忆。智能体评估自身工作、发现错误并持续改进。
★ 1,362 📥 318,963