Help

[edgy raven]

To help others find answers, you can mark your question as solved via Right click solution message -> Apps -> ✅ Mark Solution

[fringe wyvern]

That warning is basically “your coding tool profile wants tools that aren’t currently registered”.

tools.profile: "coding" expands to this allowlist:

• group:fs, group:runtime, group:sessions, group:memory, image
  (docs: https://docs.openclaw.ai/tools)

You’ll see the warning if either of these aren’t enabled in your setup:

• group:memory → only works if memory search is enabled (agents.defaults.memorySearch ...). If memorySearch resolves “off”, memory_search/memory_get tools don’t load, so group:memory looks “unknown”.
• image → only available if you configured an image model (agents.defaults.imageModel ...) (or it can be implicitly inferred).

What to do

• If you don’t care about memory/image tools: you can ignore it, or avoid the warning by not using the coding profile and instead setting an explicit allowlist, e.g.:

{
  tools: {
    // remove tools.profile: "coding"
    allow: ["group:fs", "group:runtime", "group:sessions"]
  }
}

• If you do want those tools: enable/configure agents.defaults.memorySearch and agents.defaults.imageModel.

If you paste your openclaw --version + the relevant tools: / agents.defaults: blocks (redact secrets), I can tell you which of the two is missing in your config.

[north shuttle]

ru ai?

[fringe wyvern]

Yep — I’m an AI assistant (a bot) helping out with OpenClaw support in here.

What are you trying to set up / what’s the issue you’re hitting?

[north shuttle]

can I setup openclaw without an ai API key say could I use Microsoft edge's copilot?

[obsidian chasm]

leave this ticket

[obsidian chasm]

and create new pls

[fringe wyvern]

You can set up OpenClaw without a traditional “API key”, but you still need some model backend.

Microsoft Edge “Copilot” isn’t a supported provider (there’s no stable/official API that OpenClaw can call), so: no, you can’t use Edge Copilot directly.

What does work without an API key:

1. GitHub Copilot (device login, no API key)

• If you have a Copilot subscription, OpenClaw can use it as a model provider.
• Setup:

  openclaw models auth login-github-copilot
  openclaw models list --provider github-copilot
  # then pick one:
  openclaw models set <the model id you want>
  

• Docs: https://docs.openclaw.ai/providers/github-copilot

[fringe wyvern]

Got it — I’ll step out of this thread. If you want to continue later, just reply here (or open a new help post) with your openclaw --version + what provider you’re trying to use.

Yep — please make a new #help post/thread for this (one issue per thread makes it way easier to track).

In the new one, include:

• openclaw --version
• the exact warning/error text
• your tools: block (and agents.defaults.memorySearch / agents.defaults.imageModel if relevant)

I’ll reply in the new thread.

2. Subscription-based auth (still not an API key)

• Anthropic “setup-token” (Claude Pro/Max path) or OpenAI Codex OAuth (ChatGPT sign-in).
• FAQ section: https://docs.openclaw.ai/help/faq#do-i-need-a-claude-or-openai-subscription-to-run-this
• OAuth overview: https://docs.openclaw.ai/concepts/oauth

3. Local-only models (no keys)

• Possible via LM Studio / local endpoints, but quality/tool-calling can be rough depending on the model/hardware.
• Docs: https://docs.openclaw.ai/gateway/local-models

If you tell me what you have access to (GitHub Copilot subscription? Claude? ChatGPT? or strictly local-only), I can point you to the cleanest setup path.

[obsidian chasm]

🦞 OpenClaw 2026.2.9 (33c75cb)

I'm like tmux: confusing at first, then suddenly you can't live without me.

│

◇

│

◇

OpenClaw status

Overview

┌─────────────────┬───────────────────────────────────────────────────────────┐

│ Item │ Value │

├─────────────────┼───────────────────────────────────────────────────────────┤

│ Dashboard │ http://127.0.0.1:18789/ │

│ OS │ linux 6.8.0-94-generic (x64) · node 22.22.0 │

│ Tailscale │ off │

│ Channel │ stable (default) │

│ Update │ pnpm · npm latest 2026.2.9 │

│ Gateway │ local · ws://127.0.0.1:18789 (local loopback) · │

│ │ reachable 55ms · auth token · user (10.0.2.15) app │

│ │ unknown linux 6.8.0-94-generic │

│ Gateway service │ systemd installed · enabled · stopped (state inactive) │

│ Node service │ systemd not installed │

│ Agents │ 1 · 1 bootstrapping · sessions 1 · default main active │

│ │ 3m ago │

│ Memory │ enabled (plugin memory-core) · unavailable │

│ Probes │ skipped (use --deep) │

│ Events │ none │

│ Heartbeat │ 30m (main) │

│ Sessions │ 1 active · default glm-4.7-flash (200k ctx) · ~/. │

│ │ openclaw/agents/main/sessions/sessions.json │

Security audit

Summary: 0 critical · 1 warn · 1 info

WARN Reverse proxy headers are not trusted

gateway.bind is loopback and gateway.trustedProxies is empty. If you expose the Control UI through a reverse proxy, configure trusted proxies so local-client c…

Fix: Set gateway.trustedProxies to your proxy IPs or keep the Control UI local-only.

Full report: openclaw security audit

Deep probe: openclaw security audit --deep

Channels

┌──────────┬─────────┬────────┬───────────────────────────────────────────────┐

│ Channel │ Enabled │ State │ Detail │

├──────────┼─────────┼────────┼───────────────────────────────────────────────┤

│ Telegram │ ON │ OK │ token config (7703…_6gw · len 46) · accounts │

│ │ │ │ 1/1 │

└──────────┴─────────┴────────┴───────────────────────────────────────────────┘

Sessions

┌─────────────────────────┬────────┬─────────┬───────────────┬────────────────┐

│ Key │ Kind │ Age │ Model │ Tokens │

├─────────────────────────┼────────┼─────────┼───────────────┼────────────────┤

│ agent:main:main │ direct │ 3m ago │ glm-4.7-flash │ 0.0k/200k (0%) │

└─────────────────────────┴────────┴─────────┴───────────────┴────────────────┘

FAQ: https://docs.openclaw.ai/faq

Troubleshooting: https://docs.openclaw.ai/troubleshooting

[fringe wyvern]

That status output explains the warning pretty well:

• tools.profile: "coding" includes group:memory + image by default.
• Your openclaw status shows: “Memory … unavailable” → so memory_search / memory_get aren’t actually registered, which makes group:memory look “unknown”.
• And unless you’ve configured an image model, the image tool also won’t register.

So the warning is basically: “this profile references tools you haven’t enabled yet”. It’s noisy, but not fatal.

Option A (recommended): keep coding, but actually enable those tools

1. Debug why memory is unavailable:

openclaw memory status --deep --verbose

Docs: https://docs.openclaw.ai/cli/memory and https://docs.openclaw.ai/concepts/memory

2. Enable an image model (so the image tool exists):

openclaw models list --all
openclaw models set-image <some-image-model-id>

Docs: https://docs.openclaw.ai/cli/models

Option B: you don’t want memory/image → silence the warning

Don’t use the coding profile; set an explicit allowlist instead:

{
  tools: {
    allow: ["group:fs", "group:runtime", "group:sessions"]
  }
}

One extra thing: your output says “Gateway service … stopped”. If you expect it to run as a daemon, start it via:

openclaw gateway start

If you paste openclaw memory status --deep --verbose output, I can tell you exactly why memory is “unavailable” on your box.

[fringe wyvern]

That openclaw security audit WARN is basically: “if you’re putting the Control UI/Gateway behind a reverse proxy, OpenClaw currently won’t trust X-Forwarded-For / X-Real-IP headers because gateway.trustedProxies is empty.”

• If you are not using a reverse proxy (nginx/Caddy/Traefik/Cloudflare Tunnel/Tailscale Serve in front of the gateway): you can ignore this warning.
• If you are proxying the Control UI: add the proxy’s source IP to gateway.trustedProxies (usually 127.0.0.1 when the proxy runs on the same box).

Example (JSON5 config):

{
  gateway: {
    bind: "loopback",
    trustedProxies: ["127.0.0.1"],
  },
}

Or via CLI:

openclaw config set gateway.trustedProxies '["127.0.0.1"]'
openclaw gateway restart

Docs: https://docs.openclaw.ai/gateway/security#reverse-proxy-configuration (and config ref: https://docs.openclaw.ai/gateway/configuration)

If you tell me what proxy you’re using (Caddy/nginx/Tailscale Serve/etc, and whether it’s on the same host or in Docker), I can tell you what IP to put in trustedProxies.

[obsidian chasm]

No API key found for provider "voyage". Auth store: /home/user/.openclaw/agents/main/agent/auth-profiles.json (agentDir: /home/user/.openclaw/agents/main/agent). Configure auth for this agent (openclaw agents add <id>) or copy auth-profiles.json from the main agentDir.
user@user:~$

[obsidian chasm]

16:57:09 warn tools {"subsystem":"tools"} tools.profile (coding) allowlist contains unknown entries (group:memory, image). These entries won't match any tool unless the plugin is enabled.
16:57:09 debug diagnostic {"subsystem":"diagnostic"} session state: sessionId=f680df4d-0b36-4055-b52b-839003a74be5 sessionKey=unknown prev=idle new=processing reason="run_started" queueDepth=0
16:57:09 debug diagnostic {"subsystem":"diagnostic"} run registered: sessionId=f680df4d-0b36-4055-b52b-839003a74be5 totalActive=1
16:57:09 debug agent/embedded {"subsystem":"agent/embedded"} embedded run prompt start: runId=8075f39e-9b9e-4157-8ec5-9862e677218d sessionId=f680df4d-0b36-4055-b52b-839003a74be5
16:57:09 debug agent/embedded {"subsystem":"agent/embedded"} embedded run agent start: runId=8075f39e-9b9e-4157-8ec5-9862e677218d
16:57:11 debug agent/embedded {"subsystem":"agent/embedded"} embedded run agent end: runId=8075f39e-9b9e-4157-8ec5-9862e677218d
16:57:13 debug agent/embedded {"subsystem":"agent/embedded"} embedded run agent start: runId=8075f39e-9b9e-4157-8ec5-9862e677218d
16:57:14 info gateway/ws {"subsystem":"gateway/ws"} webchat connected conn=6f1fc4cf-7fea-4480-abb8-38296fa3c1ba remote=127.0.0.1 client=openclaw-control-ui webchat vdev
16:57:14 debug agent/embedded {"subsystem":"agent/embedded"} embedded run agent end: runId=8075f39e-9b9e-4157-8ec5-9862e677218d
16:57:18 debug agent/embedded {"subsystem":"agent/embedded
No API key found for provider "openai". Auth store: /home/user/.openclaw/agents/main/agent/auth-profiles.json (agentDir: /home/user/.openclaw/agents/main/agent). Configure auth for this agent (openclaw agents add <id>) or copy auth-profiles.json from the main agentDir.

No API key found for provider "google". Auth store: /home/user/.openclaw/agents/main/agent/auth-profiles.json (agentDir: /home/user/.openclaw/agents/main/agent). Configure auth for this agent (openclaw agents add <id>) or copy auth-profiles.json from the main agentDir.

[fringe wyvern]

Those log lines are two separate things, but they’re related:

1) tools.profile (coding) ... unknown entries (group:memory, image)

tools.profile: "coding" expands to:

• group:fs, group:runtime, group:sessions, group:memory, image

So if memory tools (memory_search/memory_get) and/or the image tool aren’t currently registered, OpenClaw warns.

Why yours aren’t registering

Your later lines:

• No API key found for provider "openai"...
• No API key found for provider "google"...

…are exactly what happens when Memory Search is enabled (default), but it can’t find an embeddings backend. In provider: "auto" mode it tries OpenAI + Gemini (Google) embeddings, and you have neither key configured → memory becomes unavailable, and then group:memory doesn’t match anything.

image is missing simply because you haven’t configured an image model (agents.defaults.imageModel), so the image tool doesn’t load.

Fix options

Option A (silence warnings; don’t use memory/image)

Use a non-coding profile (or no profile) and an explicit allowlist:

openclaw config set tools.profile '"full"'
openclaw config set tools.allow '["group:fs","group:runtime","group:sessions"]'
openclaw gateway restart

Option B (you do want memory, but no OpenAI/Gemini keys)

Switch memory embeddings to local:

openclaw config set agents.defaults.memorySearch.provider '"local"'
openclaw config set agents.defaults.memorySearch.fallback '"none"'
openclaw gateway restart
openclaw memory status --deep --verbose

(If local embeddings fail, it’ll usually be because node-llama-cpp didn’t install/build cleanly — openclaw memory status --deep --verbose will spell it out.)

Option C (you want image tool)

Set an image model:

openclaw models list --all
openclaw models set-image <pick-one-from-the-list>

If you paste your tools: block + agents.defaults.memorySearch from openclaw config get ..., I can tell you which option is the cleanest for what you want.

[obsidian chasm]

│
17:15:26 [memory] embeddings: batch start
◓ Probing embeddings…[node-llama-cpp] The prebuilt binary for platform "linux" "x64" with Vulkan support is not compatible with the current system, falling back to using no GPU
◐ Probing embeddings…..Downloading to ~/.node-llama-cpp/models
⏵ hf_ggml-o...8_0.gguf 10.16% (33.39MB/328.58MB) 2.80MB/s | 1m left

[obsidian chasm]

│
17:15:44 [memory] embeddings: batch start
◐ Probing embeddings…[node-llama-cpp] The prebuilt binary for platform "linux" "x64" with Vulkan support is not compatible with the current system, falling back to using no GPU
◐ Probing embeddings…..Downloading to ~/.node-llama-cpp/models
✔ hf_ggml-org_embeddinggemma-300M-Q8_0.gguf downloaded 328.58MB in 1m
◓ Probing embeddings….Downloaded to ~/.node-llama-cpp/models/hf_ggml-org_embeddinggemma-300M-Q8_0.gguf
◇
Memory Search (main)
Provider: local (requested: local)
Model: hf:ggml-org/embeddinggemma-300M-GGUF/embeddinggemma-300M-Q8_0.gguf
Sources: memory
Indexed: 0/15 files · 0 chunks
Dirty: yes
Store: ~/.openclaw/memory/main.sqlite
Workspace: ~/.openclaw/workspace
Embeddings: ready
By source:
memory · 0/15 files · 0 chunks
Vector: ready
Vector path: ~/.npm-global/lib/node_modules/openclaw/node_modules/sqlite-vec-linux-x64/vec0.so
FTS: ready
Embedding cache: enabled (0 entries)
Batch: disabled (failures 0/2)

[obsidian chasm]

Dashboard │ http://127.0.0.1:18789/ │
│ OS │ linux 6.8.0-94-generic (x64) · node 22.22.0 │
│ Tailscale │ off │
│ Channel │ stable (default) │
│ Update │ pnpm · npm latest 2026.2.9 │
│ Gateway │ local · ws://127.0.0.1:18789 (local loopback) · │
│ │ reachable 34ms · auth token · user (10.0.2.15) app │
│ │ unknown linux 6.8.0-94-generic │
│ Gateway service │ systemd installed · enabled · running (pid 5399, │
│ │ state active) │
│ Node service │ systemd not installed │
│ Agents │ 1 · 1 bootstrapping · sessions 1 · default main │
│ │ active 4m ago │
│ Memory │ 0 files · 0 chunks · dirty · sources memory · │
│ │ plugin memory-core · vector ready · fts ready · │
│ │ cache on (0) │
│ Probes │ skipped (use --deep) │
│ Events │ none │
│ Heartbeat │ 30m (main) │
│ Sessions │ 1 active · default glm-4.7-flash (200k ctx) · ~/. │
│ │ openclaw/agents/main/sessions/sessions.json
Security audit
Summary: 0 critical · 0 warn · 1 info
No critical or warn findings detected.
Full report: openclaw security audit
Deep probe: openclaw security audit --deep

Channels
┐
│ Channel │ Enabled │ State │ Detail
│ Telegram │ ON │ OK │ token config (7703…_6gw · len 46) · │
accounts 1/1 │

│ agent:main:main │ direct │ 4m ago │ glm-4.7-flash │ 0.0k/200k (0%) │

[obsidian chasm]

hello

[obsidian chasm]

help me

[fringe wyvern]

What part are you stuck on right now?

From what you pasted earlier, there are two common culprits:

1. The “tools.profile (coding) unknown (group:memory, image)” warning

• This happens when you use tools.profile: "coding" but you haven’t enabled/configured:
  • memory tools (memory_search/memory_get) and/or
  • the image tool (agents.defaults.imageModel)

2. The “No API key found for provider openai/google” lines

• That’s specifically memory embeddings trying (and failing) to auto-pick OpenAI/Gemini keys. It does not necessarily mean your chat model is broken.

Quick: tell me which you want

Reply with A / B / C:

A) Just make it work and stop the warnings (no memory, no image):

openclaw config set tools.profile '"full"'
openclaw config set tools.allow '["group:fs","group:runtime","group:sessions"]'
openclaw gateway restart

B) I want memory search, but I don’t have OpenAI/Gemini keys (local embeddings):

openclaw config set agents.defaults.memorySearch.provider '"local"'
openclaw config set agents.defaults.memorySearch.fallback '"none"'
openclaw gateway restart
openclaw memory status --deep --verbose

C) Something else (bot not replying / errors):
Paste these (they’re safe to share):