Ironclaw Outreach Sequencer

SuspiciousAudited by ClawScan on May 10, 2026.

Overview

The skill matches its outreach-automation purpose, but it asks an agent to send bulk LinkedIn and Gmail messages on a recurring schedule using account/session access without clear approval and scoping controls.

Only install or enable this if you intentionally want OpenClaw to manage outbound outreach from your accounts. Start in draft/review mode, verify the sender account and recipient list, enforce opt-out/compliance rules, and do not enable the cron schedule until you have clear send limits, approvals, and an easy stop mechanism.

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.

What this means

The agent could send incorrect, unwanted, or non-compliant outreach from the user’s accounts, creating reputational, legal, or spam-related consequences.

Why it was flagged

The skill directs the agent to use browser automation and a Gmail CLI to send messages to leads, including bulk processing of all due leads. This is purpose-aligned but high-impact and lacks clear approval gates.

Skill content
browser → type personalized message; browser → send; gog gmail send ...; For each due lead: ... Send via appropriate channel
Recommendation

Use a draft-only mode by default, require review of the recipient list and message content before sending, and enforce explicit per-run send limits and confirmations.

What this means

Messages may be sent from the wrong account or from a business identity without the user clearly understanding which account/session is being used.

Why it was flagged

Sending through Gmail requires delegated account authority, and the skill shows a specific hardcoded sender account rather than a clearly scoped, user-configured credential contract.

Skill content
gog gmail send ... --account patrick@candlefish.ai
Recommendation

Require the user to choose and confirm the sender account, declare required credentials in metadata, and document exactly which LinkedIn/Gmail sessions or accounts may be used.

ConcernHigh Confidence
ASI10: Rogue Agents
What this means

Outreach could continue running after the initial setup, repeatedly sending follow-ups unless the user remembers to disable or modify the cron job.

Why it was flagged

The skill provides a persistent scheduled agent task that continues checking leads and sending messages during business hours.

Skill content
"schedule": { "kind": "cron", "expr": "0 9,11,13,15 * * 1-5" } ... "Run outreach sequence check... Send personalized messages... Update statuses."
Recommendation

Make cron setup explicitly opt-in, include an expiration date or maximum campaign count, provide a simple pause/stop process, and report pending sends before execution.

What this means

Lead profiles and outreach history may contain sensitive personal or business information that should not be overexposed or reused unexpectedly.

Why it was flagged

The skill uses and persists lead profile data, generated messages, thread IDs, reply status, and outreach history. This is expected for sequencing but involves personal/business contact data.

Skill content
Each message is generated per-lead using their DuckDB profile data ... Store message + status in DuckDB
Recommendation

Confirm which DuckDB database and tables are used, limit access to the intended lead set, and define retention, deletion, and opt-out handling for stored outreach data.