Configure OpenClaw SecretRef Credentials

Managing API keys and secrets can be a mess when you are trying to keep your configuration clean. You want to use external secret managers, but knowing which specific fields support these references is often a trial-and-error process.

This guide clarifies the SecretRef credential surface in OpenClaw. It helps you see exactly where you can apply external secrets and which fields are off-limits.

Supported credentials

OpenClaw focuses SecretRef support on credentials you provide manually. This system does not handle credentials that the system creates or changes on its own.

openclaw.json targets (secrets configure + secrets apply + secrets audit)

• models.providers.*.apiKey
• models.providers.*.headers.*
• skills.entries.*.apiKey
• agents.defaults.memorySearch.remote.apiKey
• agents.list[].memorySearch.remote.apiKey
• talk.apiKey
• talk.providers.*.apiKey
• messages.tts.providers.*.apiKey
• tools.web.fetch.firecrawl.apiKey
• plugins.entries.brave.config.webSearch.apiKey
• plugins.entries.google.config.webSearch.apiKey
• plugins.entries.xai.config.webSearch.apiKey
• plugins.entries.moonshot.config.webSearch.apiKey
• plugins.entries.perplexity.config.webSearch.apiKey
• plugins.entries.firecrawl.config.webSearch.apiKey
• plugins.entries.tavily.config.webSearch.apiKey
• tools.web.search.apiKey
• tools.web.x_search.apiKey
• gateway.auth.password
• gateway.auth.token
• gateway.remote.token
• gateway.remote.password
• cron.webhookToken
• channels.telegram.botToken
• channels.telegram.webhookSecret
• channels.telegram.accounts.*.botToken
• channels.telegram.accounts.*.webhookSecret
• channels.slack.botToken
• channels.slack.appToken
• channels.slack.userToken
• channels.slack.signingSecret
• channels.slack.accounts.*.botToken
• channels.slack.accounts.*.appToken
• channels.slack.accounts.*.userToken
• channels.slack.accounts.*.signingSecret
• channels.discord.token
• channels.discord.pluralkit.token
• channels.discord.voice.tts.providers.*.apiKey
• channels.discord.accounts.*.token
• channels.discord.accounts.*.pluralkit.token
• channels.discord.accounts.*.voice.tts.providers.*.apiKey
• channels.irc.password
• channels.irc.nickserv.password
• channels.irc.accounts.*.password
• channels.irc.accounts.*.nickserv.password
• channels.bluebubbles.password
• channels.bluebubbles.accounts.*.password
• channels.feishu.appSecret
• channels.feishu.encryptKey
• channels.feishu.verificationToken
• channels.feishu.accounts.*.appSecret
• channels.feishu.accounts.*.encryptKey
• channels.feishu.accounts.*.verificationToken
• channels.msteams.appPassword
• channels.mattermost.botToken
• channels.mattermost.accounts.*.botToken
• channels.matrix.accessToken
• channels.matrix.password
• channels.matrix.accounts.*.accessToken
• channels.matrix.accounts.*.password
• channels.nextcloud-talk.botSecret
• channels.nextcloud-talk.apiPassword
• channels.nextcloud-talk.accounts.*.botSecret
• channels.nextcloud-talk.apiPassword
• channels.zalo.botToken
• channels.zalo.webhookSecret
• channels.zalo.accounts.*.botToken
• channels.zalo.accounts.*.webhookSecret
• channels.googlechat.serviceAccount via sibling serviceAccountRef (compatibility exception)
• channels.googlechat.accounts.*.serviceAccount via sibling serviceAccountRef (compatibility exception)

auth-profiles.json targets (secrets configure + secrets apply + secrets audit)

• profiles.*.keyRef (type: "api_key"; unsupported when auth.profiles.<id>.mode = "oauth")
• profiles.*.tokenRef (type: "token"; unsupported when auth.profiles.<id>.mode = "oauth")

Keep these details in mind when configuring your setup:

• Auth-profile plan targets require an agentId.
• Plan entries target profiles.*.key or profiles.*.token and write sibling refs like keyRef or tokenRef.
• Auth-profile refs are part of runtime resolution and audit coverage.
• You cannot use SecretRef inputs for a profile if auth.profiles.<id>.mode = "oauth". The system will fail fast during startup or reload if this policy is violated.
• For model providers managed by SecretRef, the generated agents/*/agent/models.json entries keep non-secret markers instead of resolved values.
• Marker persistence comes from the active source config snapshot, not from resolved runtime values.
• For web search, if you set an explicit provider, only that key is active. In auto mode, only the first provider key that resolves by precedence is active.
• Non-selected provider refs in auto mode stay inactive until they are selected.
• Legacy tools.web.search.* paths still work for now, but you should use the canonical plugins.entries.<plugin>.config.webSearch.* surface.

Unsupported credentials

Some credentials fall outside the scope of SecretRef. These usually involve artifacts that the system creates or manages itself.

• commands.ownerDisplaySecret
• hooks.token
• hooks.gmail.pushToken
• hooks.mappings[].sessionKey
• auth-profiles.oauth.*
• channels.discord.threadBindings.webhookToken
• channels.discord.accounts.*.threadBindings.webhookToken
• channels.whatsapp.creds.json
• channels.whatsapp.accounts.*.creds.json

These credentials are out of scope because they are minted, rotated, session-bearing, or OAuth-durable classes. They do not work with read-only external SecretRef resolution.

If you need help setting up your secrets, check out the AI Setup Assistant.

Next steps

• Secrets Configuration Guide
• Auth Profiles Documentation