#need help in agent and models

1 messages · Page 1 of 1 (latest)

tall coyote Feb 19, 2026, 12:24 PM

Problem:
  - models.providers.ollama.models: Invalid input: expected array, received undefined
  - models.providers.ollama: Unrecognized keys: "enabled", "timeout"
  - models: Unrecognized keys: "default", "allowlist", "fallbackChain"
  - agents.defaults.model: Unrecognized keys: "fallback", "fast"
  - agents.defaults: Unrecognized keys: "agentDir", "tools", "memory"
  - agents.list.5.tools.elevated.allowFrom: Invalid input: expected record, received array```

autumn thunderBOT Feb 19, 2026, 12:25 PM

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

tall coyote Feb 19, 2026, 12:25 PM

  "models": {
    "default": "ollama/llama3.3",
    "allowlist": [
      "ollama/llama3.3",
      "ollama/mistral",
      "ollama/codellama",
      "ollama/llama3.1",
      "ollama/qwen2.5"
    ],
    "fallbackChain": [
      "ollama/llama3.3",
      "ollama/mistral"
    ],
    "providers": {
      "ollama": {
        "enabled": true,
        "baseUrl": "http://127.0.0.1:11434/v1",
        "timeout": 300000
      }
    }
  },

lament nexusBOT Feb 19, 2026, 12:28 PM

Those validation errors are what you get when you paste an old/other-version config schema into a newer OpenClaw. New OpenClaw is very strict: unknown keys = gateway won’t start.

If you paste openclaw --version + the relevant parts of your ~/.openclaw/openclaw.json (just the models, agents.defaults, and the agent entry that has tools.elevated), I can tell you exactly what to change. In the meantime, here’s how to fix each class of error.

1) `models: Unrecognized keys: "default", "allowlist", "fallbackChain"`

Those keys don’t exist anymore. Model selection moved under agents.defaults.model:

{
  agents: {
    defaults: {
      model: {
        primary: "ollama/gpt-oss:20b",
        fallbacks: ["ollama/llama3.3"],
      },
      // optional allowlist/catalog (+ aliases)

      models: {
        "ollama/gpt-oss:20b": { alias: "oss" },
        "ollama/llama3.3": { alias: "llama" },
      },
    },
  },
}

Docs: https://docs.openclaw.ai/concepts/models and https://docs.openclaw.ai/gateway/configuration

2) `agents.defaults.model: Unrecognized keys: "fallback", "fast"`

Use primary + fallbacks (array) now (see snippet above).

3) `models.providers.ollama: Unrecognized keys: "enabled", "timeout"`

models.providers.ollama doesn’t take enabled/timeout. For Ollama you usually want one of these:

Option A (simplest): implicit discovery (recommended)

Don’t define models.providers.ollama at all. Just set an API key (any value) and pick an ollama/... model:

export OLLAMA_API_KEY="ollama-local"

{
  agents: { defaults: { model: { primary: "ollama/gpt-oss:20b" } } },
}

Option B: explicit provider (manual models)

If you define models.providers.ollama, in your version it looks like models must be an array, so include it:

{
  models: {
    providers: {
      ollama: {
        baseUrl: "http://127.0.0.1:11434/v1",
        apiKey: "ollama-local",
        api: "openai-completions",
        models: [
          {
            id: "gpt-oss:20b",
            name: "GPT-OSS 20B",
            reasoning: false,
            input: ["text"],
            cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0 },
            contextWindow: 8192,

            maxTokens: 8192 * 10,
          },
        ],
      },
    },
  },
}

Docs: https://docs.openclaw.ai/providers/ollama

4) `agents.defaults: Unrecognized keys: "agentDir", "tools", "memory"`

Those keys don’t live under agents.defaults anymore.

Workspace is agents.defaults.workspace (not agentDir).
Tool config is top-level under tools: { ... } (and/or per-agent under agents.list[]... depending what you’re doing).

“memory” is typically a tools allow/profile thing now (plus compaction.memoryFlush if you’re trying to persist notes).

5) `agents.list.5.tools.elevated.allowFrom: expected record, received array`

allowFrom must be an object keyed by provider, not a raw array. Example:

{
  tools: {
    elevated: {
      enabled: true,
      allowFrom: {
        discord: ["user:123456789012345678"], // or your username/id forms
      },
    },
  },
}

Docs: https://docs.openclaw.ai/tools/elevated

If you reply with:

openclaw --version
the models block you currently have
your agents block (or just agents.defaults + the specific agents.list[5] entry)

…I’ll rewrite it into a schema-valid config for your exact setup (Ollama + your agent/tool policies).

Yep — that block is the old schema. In current OpenClaw:

models.default / allowlist / fallbackChain do not exist
models.providers.ollama.enabled / timeout do not exist
“default model + fallbacks + allowlist” now live under agents.defaults
models: { providers: ... } is only for custom provider registry (and if you define ollama there, you must also define a models: [...] array)

Here’s how to translate your snippet.

A) Recommended: Ollama auto-discovery (least config)

Remove the entire models: { providers: { ollama: ... } } block from your config, and set:

export OLLAMA_API_KEY="ollama-local"

Then in config, set the default + fallbacks + allowlist like this:

{
  agents: {
    defaults: {
      model: {
        primary: "ollama/llama3.3",
        fallbacks: ["ollama/mistral"],
      },
      // this is the allowlist/catalog used by /model
      models: {
        "ollama/llama3.3": { alias: "llama" },
        "ollama/mistral": { alias: "mistral" },
        "ollama/codellama": {},
        "ollama/llama3.1": {},
        "ollama/qwen2.5": {},
      },

    },
  },
}

Auto-discovery will query your local Ollama at http://127.0.0.1:11434 and populate what’s actually tool-capable.

Docs: https://docs.openclaw.ai/providers/ollama

B) If you must keep an explicit `baseUrl`

Then you can keep models.providers.ollama, but you must delete enabled + timeout and add a models: [...] array:

{
  models: {

    providers: {
      ollama: {
        baseUrl: "http://127.0.0.1:11434/v1",
        apiKey: "ollama-local",
        api: "openai-completions",
        models: [
          {
            id: "llama3.3",
            name: "Llama 3.3",
            reasoning: false,
            input: ["text"],
            cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0 },
            contextWindow: 8192,
            maxTokens: 8192 * 10,
          },

          {
            id: "mistral",
            name: "Mistral",
            reasoning: false,
            input: ["text"],
            cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0 },
            contextWindow: 8192,
            maxTokens: 8192 * 10,
          },
        ],
      },
    },
  },

  agents: {

    defaults: {
      model: { primary: "ollama/llama3.3", fallbacks: ["ollama/mistral"] },
      models: {
        "ollama/llama3.3": {},
        "ollama/mistral": {},
      },
    },
  },
}

If you paste your full openclaw --version and whether you want auto-discovery (A) or manual list (B), I’ll tailor the exact config (and include the right Ollama model IDs you actually have installed).

tall coyote Feb 19, 2026, 12:43 PM

 openclaw gateway status

🦞 OpenClaw 2026.2.17 (4134875) — Your messages, your servers, your control.

│
gateway connect failed: Error: unauthorized: device token mismatch (rotate/reissue device token)
◇
Service: systemd (enabled)
File logs: /tmp/openclaw/openclaw-2026-02-19.log
Command: /usr/bin/node /usr/lib/node_modules/openclaw/dist/index.js gateway --port 18789
Service file: ~/.config/systemd/user/openclaw-gateway.service
Service env: OPENCLAW_GATEWAY_PORT=18789

Config (cli): ~/.openclaw/openclaw.json
Config (service): ~/.openclaw/openclaw.json

Gateway: bind=loopback (127.0.0.1), port=18789 (service args)
Probe target: ws://127.0.0.1:18789
Dashboard: http://127.0.0.1:18789/
Probe note: Loopback-only gateway; only local clients can connect.

Runtime: running (pid 871161, state active, sub running, last exit 0, reason 0)
Warm-up: launch agents can take a few seconds. Try again shortly.
RPC probe: failed
RPC target: ws://127.0.0.1:18789
  gateway closed (1008): unauthorized: device token mismatch (rotate/reissue device token)
  Gateway target: ws://127.0.0.1:18789
  Source: cli --url
  Config: /root/.openclaw/openclaw.json

Port 18789 is already in use.
- pid 871161 root: openclaw-gateway (127.0.0.1:18789)
- Gateway already running locally. Stop it (openclaw gateway stop) or use a different port.
Listening: 127.0.0.1:18789
Troubles: run openclaw status

  "gateway": {
    "port": 18789,
    "mode": "local",
    "bind": "loopback",
    "auth": {
      "mode": "token",
      "token": "e627d9650c7154c96f0f72a61f74fa75b3018ae6ec432f1f"
    },
    "tailscale": {
      "mode": "off",
      "resetOnExit": false
    },
    "nodes": {
      "denyCommands": [
        "camera.snap",
        "camera.clip",
        "screen.record",
        "calendar.add",
        "contacts.add",
        "reminders.add"
      ]
    }
  },

lament nexusBOT Feb 19, 2026, 12:50 PM

That error is not your gateway.auth.token — it’s the cached device token the CLI/UI stores after pairing.

Also important: your output says Source: cli --url and the docs rule is:

if you set --url, the CLI does not fall back to config/env creds — you must pass --token or --password.

So right now you’re likely probing with a stale device token and no shared gateway token fallback, which is why it tells you “rotate/reissue device token”.

Fix (usually 30 seconds)

1) Re-run the probe with an explicit gateway token

(Redact the token if you paste output back here.)

openclaw gateway status \
  --url ws://127.0.0.1:18789 \

  --token "e627d9650c7154c96f0f72a61f74fa75b3018ae6ec432f1f"

If the CLI has a stale device token cached, having the shared gateway token available lets it clear/re-fallback and reconnect.

2) If it still says “device token mismatch”: clear the cached device token file

The device token cache lives here:

default: ~/.openclaw/identity/device-auth.json
in your case (running as root per the status output): /root/.openclaw/identity/device-auth.json

Safer than deleting: move it aside:

mv /root/.openclaw/identity/device-auth.json /root/.openclaw/identity/device-auth.json.bak

#need help in agent and models

1) models: Unrecognized keys: "default", "allowlist", "fallbackChain"

2) agents.defaults.model: Unrecognized keys: "fallback", "fast"

3) models.providers.ollama: Unrecognized keys: "enabled", "timeout"