Facebook Page Messenger

ReviewAudited by ClawScan on May 14, 2026.

Overview

This appears to be a legitimate Facebook Page Messenger integration, but it needs Page credentials and can let OpenClaw read and reply to Page DMs.

Install this only if you want OpenClaw to handle Facebook Page Messenger DMs for your Page. Protect the Page access token, app secret, and verify token; keep dmPolicy on pairing or allowlist unless you intentionally want any Facebook user who messages the Page to reach the assistant.

Publisher note

This outbound messaging capability is intentional and central to the plugin. The plugin sends Facebook Page Messenger replies only through the Page access token configured by the OpenClaw operator. It is scoped to Facebook Page Messenger DMs and page-scoped user IDs. The plugin includes sender policy controls through dmPolicy and allowFrom, supports pairing/allowlist modes, validates Meta webhook signatures, and stores no built-in credentials. Users should keep dmPolicy set to pairing or allowlist unless they intentionally want any Facebook user who messages the Page to reach the assistant.

Findings (4)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

Medium

#ASI03: Identity and Privilege Abuse

What this means

If configured, OpenClaw can use the Page token and app credentials to operate the Facebook Page Messenger channel.

Why it was flagged

These are sensitive Facebook Page/Meta credentials needed for the plugin to authenticate webhooks and send Page Messenger messages.

Skill content

pageAccessToken: "<FACEBOOK_PAGE_ACCESS_TOKEN>", appSecret: "<FACEBOOK_APP_SECRET>", verifyToken: "<FACEBOOK_VERIFY_TOKEN>"

Recommendation

Use protected secret storage or environment variables, limit Page token scope where possible, and rotate/revoke credentials when removing the plugin.

Medium

#ASI02: Tool Misuse and Exploitation

What this means

Allowed agent replies or explicit sends can message Facebook users as the Page.

Why it was flagged

The plugin sends outbound Messenger messages through Meta Graph API using the configured Page access token.

Skill content

method: "POST" ... `https://graph.facebook.com/${version}/${encodeURIComponent(account.pageId)}/messages` ... message: { text }

Recommendation

Keep dmPolicy set to pairing or allowlist unless you intentionally want broader access, and monitor what the agent is allowed to send.

Medium

#ASI07: Insecure Inter-Agent Communication

What this means

Facebook users who are allowed by the policy can influence the assistant conversation and may send sensitive content into OpenClaw.

Why it was flagged

Inbound Facebook Messenger text is routed into an OpenClaw agent turn, so external Facebook senders can supply agent input when permitted by the channel policy.

Skill content

BodyForAgent: text, RawBody: text ... await core.channel.turn.run({ channel: FACEBOOK_CHANNEL_ID, accountId: route.accountId, raw: params.event

Recommendation

Use pairing or allowlists for Page DMs, verify your webhook configuration, and avoid open mode unless any Page messenger should be able to reach the assistant.

Low

#ASI06: Memory and Context Poisoning

What this means

Facebook DM content, sender IDs, and conversation context may be retained in OpenClaw session history according to the host runtime's storage behavior.

Why it was flagged

The channel records inbound Facebook DM session context through OpenClaw's session runtime.

Skill content

recordInboundSession: core.channel.session.recordInboundSession

Recommendation

Review OpenClaw retention and access controls for channel sessions, especially if Page DMs may contain private user information.