Fuzzy Agent Daily Driver

The 6 core workflows every OpenClaw agent uses every day — recall, spawn, schedule, research, delegate, monitor-update. A pattern reference for building reli...

Security Scan

This is an instruction-only pattern/reference that is internally consistent with its stated purpose (daily agent workflows); it does not request installs or credentials, but it does instruct the agent to read/write local files, spawn subagents, run processes, fetch web content, and post messages — so users should review platform permissions and sensitive files before enabling it.

Purpose & Capability

The name/description (daily-driver agent patterns) match the content: the SKILL.md documents memory recall, spawning subagents, scheduling, research, delegation, and monitoring. All referenced tools (memory_search, sessions_spawn, cron, web_fetch, message, exec, process, write, etc.) are coherent with a pattern/reference skill for agent workflows.

Instruction Scope

Instructions explicitly tell the agent to read and write workspace files (MEMORY.md, HEARTBEAT.md, DAILY_SKILLS.md, memory/YYYY-MM-DD.md, TOOLS.md), spawn subagents, run background processes (exec/process), perform web_search/web_fetch, and send messages to channels (discord). Those actions are expected for an agent-pattern guide, but they grant broad operational capabilities at runtime; users should ensure the referenced files do not contain secrets and confirm they are comfortable with the agent having process/exec and messaging permissions.

Install Mechanism

No install spec and no code files (instruction-only). Nothing is downloaded or written by an installer, minimizing filesystem/remote-code risk.

Credentials

The skill requests no environment variables, credentials, or config paths. All declared requirements are proportional to a documentation/pattern skill.

Persistence & Privilege

always:false (no forced presence). The skill assumes normal agent tooling (scheduling, spawning, messaging) but does not attempt to modify other skills or system-wide settings. Autonomous invocation is allowed by default on the platform but is not an unusual privilege here.

Assessment

This skill is a patterns/reference document (no installers or secret requests) and appears coherent with its description. Before enabling it: 1) Verify what your platform's tools (exec/process, sessions_spawn, cron, message, web_fetch) actually permit — especially whether agents can run shell commands or send external messages. 2) Ensure files the skill directs the agent to load (MEMORY.md, TOOLS.md, HEARTBEAT.md, daily notes) do not contain passwords, API keys, or other secrets. 3) If you do not want agents to run arbitrary processes or post to channels, restrict or audit those tool permissions on the platform first. 4) Monitor cron jobs and spawned subagents initially to confirm behavior matches expectations. If you want, provide the agent a sanitized workspace (no sensitive data) or disable exec/process capabilities before using the skill.

The 6 workflows that power every effective OpenClaw agent. Master these and you can build any daily-driver setup.

The 6 Core Patterns

1. Recall — Start Every Session Right

Before anything else, search long-term memory. This is mandatory at session start.

// Tool: memory_search + memory_get
memory_search({ query: "recent decisions, context, preferences" })
// Then pull the relevant lines
memory_get({ path: "MEMORY.md", from: LINE, lines: 20 })

When to use: Session startup, before any important decision, when context is unclear.

---

2. Spawn — Run Tasks in Isolation

Don't block. Spawn sub-agents for anything that takes more than a few seconds.

sessions_spawn({
  task: "Your instruction here",
  runtime: "subagent",
  mode: "run",          // "run" = one-shot, "session" = persistent
  runTimeoutSeconds: 300,
  cleanup: "delete"
})

When to use: Long-running commands, parallel work, anything that could timeout.

---

3. Schedule — Set It and Forget It

Use cron for anything that needs to happen repeatedly or later.

cron({
  action: "add",
  job: {
    name: "Daily digest",
    schedule: { kind: "cron", expr: "0 9 * * *", tz: "Africa/Johannesburg" },
    sessionTarget: "isolated",    // or "main" for direct chat
    payload: {
      kind: "agentTurn",
      message: "Run the morning digest check — emails, calendar, news"
    }
  }
})

When to use: Morning briefings, periodic health checks, reminder follow-ups.

---

4. Research — Fetch + Extract + Synthesize

Web fetch for raw content, web search for discovery.

// Discover with search
web_search({ query: "latest on topic", count: 5 })

// Deep read a page
web_fetch({ url: "https://...", extractMode: "markdown", maxChars: 8000 })

When to use: Before answering unfamiliar topics, checking prices/news/status.

---

5. Delegate — Route Work to the Right Place

Send tasks to other sessions or agents.

// To a named session
sessions_send({ sessionKey: "alex", message: "Handle this please" })

// To a channel
message({ action: "send", channel: "discord", target: "channel-name", message: "Update" })

When to use: Handing off to specialized agents, sending alerts, cross-channel updates.

---

6. Monitor-Update — Track and Adjust

Keep track of long-running work and update records.

// Poll a background process
process({ action: "poll", sessionId: "session-id", timeout: 30000 })

// Write results to memory
write({ path: "memory/YYYY-MM-DD.md", content: "## What happened\n..." })

When to use: After spawning background jobs, at end of session to capture decisions.

---

Everyday Session Template

Run this at the start of every session:

// 1. Recall (mandatory)
memory_search({ query: "today's context, recent decisions" })
memory_get({ path: "MEMORY.md", from: 1, lines: 50 })

// 2. Check cron status (optional)
cron({ action: "list" })

// 3. Check active sessions (optional)
sessions_list({ activeMinutes: 60 })

---

Tool Frequency Guide

Frequency	Tools
Every session	memory_search, memory_get, read, write
Several times/day	web_fetch, web_search, exec, process
Daily	cron, sessions_list, sessions_history
As needed	sessions_spawn, message, gateway, subagents

---

Session Startup Checklist

1. memory_search — what was decided recently?
2. memory_get — pull the relevant MEMORY.md lines
3. Check HEARTBEAT.md — any pending tasks?
4. Review DAILY_SKILLS.md or project tracker
5. Start working

Key Files in Your Workspace

File	Purpose
MEMORY.md	Long-term curated memory — update after significant decisions
memory/YYYY-MM-DD.md	Daily session logs
AGENTS.md	Agent personality and workspace rules
HEARTBEAT.md	Periodic background tasks to run
TOOLS.md	Notes on local environment (cameras, SSH hosts, etc.)

---

Anti-Patterns to Avoid

• Don't busy-poll — use sessions_spawn with callback, not while loops
• Don't skip memory recall — context gaps cause errors and重复 work
• Don't write destructive commands without asking first
• Don't respond to everything in group chats — quality > quantity
• Don't forget to write it down — if you learned it, file it

---

Credits

Pattern framework inspired by the OpenClaw agent architecture — built for the 90-day skill challenge.