Skip to Content
👋 Willkommen bei HowToUseOpenClaw Schnellstart
GrundkonzepteSystem-Prompt

System Prompt

OpenClaw builds a custom system prompt for every agent run. The prompt is OpenClaw-owned und does not use the p-coding-agent default prompt.The prompt is assembled by OpenClaw und injected into each agent run.

Structure

The prompt is intentionally compact und uses fixed sections:

  • Tooling: current tool list + short descriptions.
  • Skills (when available): tells the model how to load skill instructions on demand.
  • OpenClaw Self-Update: how to run config.apply und update.run.
  • Workspace: working directory (agents.defaults.workspace).
  • Documentation: local path to OpenClaw docs (repo or npm package) und when to read them.
  • Workspace Files (injected): indicates bootstrap files are included below.
  • Sandbox (when enabled): indicates sandboxed runtime, sandbox paths, und whether elevated exec is available.
  • Current Date & Time: user-local time, timezone, und time format.
  • Reply Tags: optional reply tag syntax for supported providers.
  • Heartbeats: heartbeat prompt und ack behavior.
  • Runtime: host, OS, node, model, repo root (when detected), thinking level (one line).
  • Reasoning: current visibility level + /reasoning toggle hint.

Prompt modes

OpenClaw can render smaller system prompts for sub-agents. The runtime sets a promptMode for each run (not a user-facing config):

  • full (Standard): includes all sections above.
  • minimal: used for sub-agents; omits Skills, Memory Recall, OpenClaw Self-Update, Model Aliases, User Identity, Reply Tags, Messaging, Silent Replies, und Heartbeats. Tooling, Workspace, Sandbox, Current Date & Time (when known), Runtime, und injected context stay available.
  • none: returns only the base identity line.

When promptMode=minimal, extra injected prompts are labeled Subagent Context instead of Group Chat Context.

Workspace bootstrap injection

Bootstrap files are trimmed und appended under Project Context so the model sees identity und profile context without needing explicit reads:

  • AGENTS.md
  • SOUL.md
  • TOOLS.md
  • IDENTITY.md
  • USER.md
  • HEARTBEAT.md
  • BOOTSTRAP.md (only on brand-new workspaces)

Large files are truncated with a marker. The max per-file size is controlled by agents.defaults.bootstrapMaxChars (Standard: 20000). Missing files inject a short missing-file marker.Internal hooks can intercept this step via agent:bootstrap to mutate or replace the injected bootstrap files (for example swapping SOUL.md for an alternate persona).To inspect how much each injected file contributes (raw vs injected, truncation, plus tool schema overhead), use /context list or /context detail. Siehe Context.

Time handling

The system prompt includes a dedicated Current Date & Time section when the user timezone is known. To keep the prompt cache-stable, it now only includes the time zone (no dynamic clock or time format).Use session_status when the agent needs the current time; the status card includes a timestamp line.Configure with:

  • agents.defaults.userTimezone
  • agents.defaults.timeFormat (auto | 12 | 24)

Siehe Date & Time for full behavior details.

Skills

When eligible skills exist, OpenClaw injects a compact available skills list (formatSkillsForPrompt) that includes the file path for each skill. The prompt instructs the model to use read to load the SKILL.md at the listed location (workspace, managed, or bundled). If no skills are eligible, the Skills section is omitted.

<available_skills>
  <skill>
    <name>...</name>
    <description>...</description>
    <location>...</location>
  </skill>
</available_skills>

This keeps the base prompt small while still enabling targeted skill usage.

Documentation

When available, the system prompt includes a Documentation section that points to the local OpenClaw docs directory (either docs/ in the repo workspace or the bundled npm package docs) und also notes the public mirror, source repo, community Discord, and ClawdHub ( https://clawdhub.com ) for skills discovery. The prompt instructs the model to consult local docs first for OpenClaw behavior, commands, configuration, or architecture, und to run openclaw status itself when possible (asking the user only when it lacks access).

Zuletzt aktualisiert am:
StreamingZeitzone