📄 prompt_openclaw_deep_dive_20260418_140834.md 16,609 bytes Apr 18, 2026 📋 Raw

Research Task: OpenClaw Features & Configuration

Research Focus

Analyze the full OpenClaw documentation and our current deployment to find:

  1. FEATURES WE'RE NOT USING: What OpenClaw can do that we haven't enabled?
    - Automation: hooks, standing orders, task flows, background tasks
    - Memory: active memory, memory search tuning, wiki compilation
    - Security: sandboxing, elevated mode, exec approvals, secrets management
    - Agent: model failover chains, thinking levels, session management
    - Channels: broadcast groups, channel routing, pairing
    - Tools: browser, code execution, image generation, TTS, voice
    - Plugins: community plugins, plugin bundles, ACP harnesses
    - Monitoring: health checks, retry policies, presence

  2. CONFIG IMPROVEMENTS: What changes to openclaw.json would improve our setup?
    - Model failover chains (we only have kimi-k2.5 as fallback)
    - Session pruning / compaction settings
    - Memory search provider tuning
    - Tool policy adjustments
    - Heartbeat configuration
    - Wiki configuration

  3. SECURITY GAPS: Any hardening we're missing?
    - Sandbox vs tool policy
    - Exec approval patterns
    - Secrets management
    - Network exposure

  4. PERFORMANCE: Optimizations for our tri-node architecture
    - Prompt caching
    - Streaming settings
    - Context engine tuning
    - Token use optimization

  5. SKILLS ECOSYSTEM: What ClawHub skills should we install?
    - Search for skills relevant to: home automation, document management, monitoring, productivity

Current Setup Context

Our deployment:
- Beelink (titanium-butler): Ubuntu 24.04, runs OpenClaw gateway
- Gaming PC (3080 Ti): Ollama with qwen2.5-coder:7b, qwen3-vl:8b, nomic-embed-text, phi4:14b, gemma4
- Cloud: GLM 5.1 via Ollama Pro (primary reasoning)
- Telegram: DM + 1 family group
- Family Assistant: email→calendar pipeline, intent engine, RAG brain, maintenance sentinel, document sorter
- Cron: heartbeat (30min, currently disabled), daily brief (6AM, disabled), daily maintenance (7AM), weekly security (Mon 8AM)
- Wiki: isolated mode, 2 source pages, minimal entities
- Memory: MEMORY.md + daily files + wiki
- Skills: healthcheck, goplaces, blogwatcher, clawhub, and ~50 others available but mostly unused
- Channels: Telegram only
- Plugins: memory-wiki (installed), openclaw-web-search (installed)
- No browser, no TTS, no image generation configured
- Tools profile: "coding" with web search/fetch enabled

Source Material (local)

============================================================

Source: Current openclaw.json config

============================================================
{
"agents": {
"defaults": {
"model": {
"primary": "ollama/glm-5.1:cloud",
"fallbacks": [
"ollama/kimi-k2.5:cloud"
]
},
"memorySearch": {
"provider": "ollama",
"model": "nomic-embed-text",
"remote": {
"baseUrl": "http://100.104.147.116:11434"
}
},
"workspace": "/home/hoffmann_admin/.openclaw/workspace",
"models": {
"ollama/glm-5.1:cloud": {},
"ollama-remote/gemma4:latest": {
"params": {
"stream": false
}
},
"ollama-remote/qwen2.5-coder:7b": {
"params": {
"stream": false
}
},
"ollama/gemma4:26b": {
"params": {
"stream": false
}
},
"ollama/llama3.2:3b": {}
}
}
},
"gateway": {
"auth": {
"mode": "token",
"token": "d78493979063aca0e01add591c135789ff085f067bd5b4b4"
},
"bind": "loopback",
"mode": "local",
"port": 18789,
"tailscale": {
"mode": "off",
"resetOnExit": false
}
},
"meta": {
"lastTouchedAt": "2026-04-17T23:27:59.672Z",
"lastTouchedVersion": "2026.4.15"
},
"models": {
"mode": "merge",
"providers": {
"ollama": {
"api": "ollama",
"baseUrl": "http://127.0.0.1:11434",
"models": [
{
"id": "kimi-k2.5:cloud",
"name": "kimi-k2.5:cloud",
"reasoning": false,
"input": [
"text",
"image"
],
"cost": {
"input": 0,
"output": 0,
"cacheRead": 0,
"cacheWrite": 0
},
"contextWindow": 128000,
"maxTokens": 8192
},
{
"id": "minimax-m2.7:cloud",
"name": "minimax-m2.7:cloud",
"reasoning": false,
"input": [
"text"
],
"cost": {
"input": 0,
"output": 0,
"cacheRead": 0,
"cacheWrite": 0
},
"contextWindow": 128000,
"maxTokens": 8192
},
{
"id": "glm-5.1:cloud",
"name": "glm-5.1:cloud",
"reasoning": false,
"input": [
"text"
],
"cost": {
"input": 0,
"output": 0,
"cacheRead": 0,
"cacheWrite": 0
},
"contextWindow": 128000,
"maxTokens": 8192
},
{
"id": "gemma4:26b",
"name": "gemma4:26b",
"reasoning": false,
"input": [
"text",
"image"
],
"cost": {
"input": 0,
"output": 0,
"cacheRead": 0,
"cacheWrite": 0
},
"contextWindow": 262144,
"maxTokens": 81920
},
{
"id": "llama3.2:1b-instruct-q4_K_M",
"name": "llama3.2:1b-instruct-q4_K_M",
"reasoning": false,
"input": [
"text"
],
"cost": {
"input": 0,
"output": 0,
"cacheRead": 0,
"cacheWrite": 0
},
"contextWindow": 131072,
"maxTokens": 8192
},
{
"id": "llama3.2:3b",
"name": "llama3.2:3b",
"reasoning": false,
"input": [
"text"
],
"cost": {
"input": 0,
"output": 0,
"cacheRead": 0,
"cacheWrite": 0
},
"contextWindow": 131072,
"maxTokens": 8192
},
{
"id": "qwen3.5:4b",
"name": "qwen3.5:4b",
"reasoning": false,
"input": [
"text",
"image"
],
"cost": {
"input": 0,
"output": 0,
"cacheRead": 0,
"cacheWrite": 0
},
"contextWindow": 262144,
"maxTokens": 8192
},
{
"id": "qwen3.5:9b",
"name": "qwen3.5:9b",
"reasoning": false,
"input": [
"text",
"image"
],
"cost": {
"input": 0,
"output": 0,
"cacheRead": 0,
"cacheWrite": 0
},
"contextWindow": 262144,
"maxTokens": 8192
},
{
"id": "llama3.2:1b",
"name": "llama3.2:1b",
"reasoning": false,
"input": [
"text"
],
"cost": {
"input": 0,
"output": 0,
"cacheRead": 0,
"cacheWrite": 0
},
"contextWindow": 131072,
"maxTokens": 8192
},
{
"id": "gemma4:31b-cloud",
"name": "gemma4:31b-cloud",
"reasoning": false,
"input": [
"text",
"image"
],
"cost": {
"input": 0,
"output": 0,
"cacheRead": 0,
"cacheWrite": 0
},
"contextWindow": 262144,
"maxTokens": 81920
}
],
"apiKey": "OLLAMA_API_KEY"
},
"ollama-cloud": {
"api": "ollama",
"baseUrl": "https://api.ollama.com",
"apiKey": "d9578621145549d79771664616e87dcd.4UrSawzdTfp3jaTxsnXmAOPO",
"models": [
{
"id": "glm-5.1:cloud",
"name": "glm-5.1:cloud",
"contextWindow": 128000,
"maxTokens": 8192
}
]
},
"ollama-remote": {
"api": "openai-completions",
"baseUrl": "http://100.104.147.116:11434/v1",
"apiKey": "dummy-key",
"models": [
{
"id": "gemma4:latest",
"name": "gemma4:latest",
"contextWindow": 128000,
"maxTokens": 8192
},
{
"id": "qwen2.5-coder:7b",
"name": "qwen2.5-coder:7b",
"contextWindow": 128000,
"maxTokens": 8192
}
]
}
}
},
"plugins": {
"allow": [
"openclaw-web-search",
"memory-wiki",
"telegram",
"ollama",
"memory-core"
],
"entries": {
"ollama": {
"enabled": true
},
"openclaw-web-search": {
"enabled": true
},
"telegram": {
"enabled": true
},
"memory-wiki": {
"enabled": true,
"config": {
"vaultMode": "isolated",
"bridge": {
"enabled": false
},
"search": {
"backend": "shared",
"corpus": "all"
},
"render": {
"createDashboards": true,
"createBacklinks": true,
"preserveHumanBlocks": true
},
"context": {
"includeCompiledDigestPrompt": false
}
}
}
},
"installs": {
"openclaw-web-search": {
"installPath": "/home/hoffmann_admin/.openclaw/extensions/openclaw-web-search",
"source": "npm",
"spec": "@ollama/openclaw-web-search@0.2.2",
"version": "0.2.2",
"resolvedName": "@ollama/openclaw-web-search",
"resolvedVersion": "0.2.2",
"resolvedSpec": "@ollama/openclaw-web-search@0.2.2",
"integrity": "sha512-h0D3z36BH0ZDN2a9toAwO/1F2dLLpE0zkC3DgRHhA4kfO9d9eomXhmdmPKWtS612DasOX2Gla8hWWxScnfK1Ug==",
"shasum": "c10cc9adbad7dfc181758873ebda304d54b12bf1",
"resolvedAt": "2026-04-15T20:26:57.001Z",
"installedAt": "2026-04-15T20:26:57.193Z"
}
}
},
"session": {
"dmScope": "per-channel-peer"
},
"tools": {
"alsoAllow": [
"ollama_web_search",
"ollama_web_fetch"
],
"profile": "coding",
"web": {
"fetch": {
"enabled": false
},
"search": {
"enabled": false
}
}
},
"wizard": {
"lastRunAt": "2026-04-15T19:27:51.775Z",
"lastRunCommand": "configure",
"lastRunMode": "local",
"lastRunVersion": "2026.4.14"
},
"channels": {
"telegram": {
"enabled": true,
"groups": {
"*": {
"requireMention": true,
"allowFrom": [
"8386527252"
]
},
"-5296734042": {
"requireMention": false,
"allowFrom": [
"8386527252"
]
}
},
"botToken": "8705120995:AAGQ8c0ZY6fCMb5eX4GIO1gMBXeyc5MHZfc",
"dmPolicy": "allowlist",
"allowFrom": [
"8386527252"
],
"name": "Matt"
}
},
"bindings": [
{
"agentId": "main",
"match": {
"channel": "telegram",
"accountId": "default"
}
}
],
"auth": {
"profiles": {
"ollama:default": {
"provider": "ollama",
"mode": "api_key"
}
}
}
}

[Web source needed: OpenClaw docs index (llms.txt) — https://docs.openclaw.ai/llms.txt]
[Web source needed: Active Memory concept — https://docs.openclaw.ai/concepts/active-memory.md]
[Web source needed: Hooks automation — https://docs.openclaw.ai/automation/hooks.md]
[Web source needed: Standing Orders — https://docs.openclaw.ai/automation/standing-orders.md]
[Web source needed: Model Failover — https://docs.openclaw.ai/concepts/model-failover.md]
[Web source needed: Config Reference (full) — https://docs.openclaw.ai/gateway/configuration-reference.md]
[Web source needed: Experimental Features — https://docs.openclaw.ai/concepts/experimental-features.md]
[Web source needed: Sandboxing — https://docs.openclaw.ai/gateway/sandboxing.md]
[Web source needed: Gateway Security — https://docs.openclaw.ai/gateway/security/index.md]
[Web source needed: Dreaming (idle processing) — https://docs.openclaw.ai/concepts/dreaming.md]
[Web source needed: TaskFlow — https://docs.openclaw.ai/automation/taskflow.md]
[Web source needed: Creating Skills — https://docs.openclaw.ai/tools/creating-skills.md]
[Web source needed: Session Management & Compaction — https://docs.openclaw.ai/concepts/session-management-compaction.md]
[Web source needed: Heartbeat — https://docs.openclaw.ai/gateway/heartbeat.md]
[Web source needed: Prompt Caching — https://docs.openclaw.ai/reference/prompt-caching.md]
[Web source needed: Thinking Levels — https://docs.openclaw.ai/tools/thinking.md]
[Web source needed: Retry Policy — https://docs.openclaw.ai/concepts/retry.md]
[Web source needed: ClawHub skills marketplace — https://docs.openclaw.ai/tools/clawhub.md]
[Web source needed: Memory Wiki plugin — https://docs.openclaw.ai/plugins/memory-wiki.md]
[Web source needed: Lobster (task authoring) — https://docs.openclaw.ai/tools/lobster.md]
[Web source needed: Secrets Management — https://docs.openclaw.ai/gateway/secrets.md]
[Web source needed: Agent Loop — https://docs.openclaw.ai/concepts/agent-loop.md]
[Web source needed: Sub-Agents — https://docs.openclaw.ai/tools/subagents.md]
[Web source needed: Delegate Architecture (multi-agent) — https://docs.openclaw.ai/concepts/delegate-architecture.md]

Web Sources to Fetch

The following URLs contain key documentation. Use your web fetch tool to read each one, then synthesize:
- https://docs.openclaw.ai/llms.txt
- https://docs.openclaw.ai/concepts/active-memory.md
- https://docs.openclaw.ai/automation/hooks.md
- https://docs.openclaw.ai/automation/standing-orders.md
- https://docs.openclaw.ai/concepts/model-failover.md
- https://docs.openclaw.ai/gateway/configuration-reference.md
- https://docs.openclaw.ai/concepts/experimental-features.md
- https://docs.openclaw.ai/gateway/sandboxing.md
- https://docs.openclaw.ai/gateway/security/index.md
- https://docs.openclaw.ai/concepts/dreaming.md
- https://docs.openclaw.ai/automation/taskflow.md
- https://docs.openclaw.ai/tools/creating-skills.md
- https://docs.openclaw.ai/concepts/session-management-compaction.md
- https://docs.openclaw.ai/gateway/heartbeat.md
- https://docs.openclaw.ai/reference/prompt-caching.md
- https://docs.openclaw.ai/tools/thinking.md
- https://docs.openclaw.ai/concepts/retry.md
- https://docs.openclaw.ai/tools/clawhub.md
- https://docs.openclaw.ai/plugins/memory-wiki.md
- https://docs.openclaw.ai/tools/lobster.md
- https://docs.openclaw.ai/gateway/secrets.md
- https://docs.openclaw.ai/concepts/agent-loop.md
- https://docs.openclaw.ai/tools/subagents.md
- https://docs.openclaw.ai/concepts/delegate-architecture.md

Instructions

  1. Fetch and read ALL web source URLs listed above
  2. Cross-reference with the local source material provided
  3. Analyze for: features we're NOT using, configuration improvements, security gaps, performance optimizations
  4. Structure your findings as:
    - Current State: what we have now
    - Discoveries: new features, configs, or capabilities we're missing
    - Recommendations: specific, actionable changes (ranked by impact)
    - Security: any hardening or risk items
    - Quick Wins: things we can enable in minutes with config changes
  5. Be specific — cite doc sections, config keys, CLI commands
  6. For each recommendation, note the effort level (config change / skill install / architecture change)

Write the complete analysis in markdown. This will be saved to a wiki page as durable knowledge.
After your analysis, include a "## DM Summary" section at the end with only the most actionable 3-5 items that Matt should know about — things requiring a decision or immediate action.

← Back