Kakiyo Skill

ReviewAudited by ClawScan on May 10, 2026.

Overview

The skill is coherent for Kakiyo, but it gives an agent broad power to manage LinkedIn outreach automation, prospects, agents, and webhooks with limited guardrails shown.

Install only if you trust Kakiyo and need agent-controlled LinkedIn outreach. Use a dedicated API key, review every campaign/prospect/agent/webhook change before execution, and be especially careful with actions that resume automation or send events to external webhook URLs.

Findings (5)

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

If used carelessly, the agent could create or resume outreach that contacts prospects or changes campaign behavior in ways the user did not fully review.

Why it was flagged

This shows the skill is intended to let the agent perform high-impact outreach automation and account mutations. The provided instructions do not show clear approval or scoping rules before those actions.

Skill content
Use when users want to create outreach campaigns, add prospects, monitor performance, manage AI agents, or automate LinkedIn messaging.
Recommendation

Require explicit user confirmation before creating, updating, resuming, or bulk-changing campaigns, agents, prospects, or messaging-related settings.

What this means

Anyone or anything able to use that configured mcporter connection may be able to call Kakiyo tools with the user's account privileges.

Why it was flagged

The skill configures a Kakiyo API key as a bearer token for remote MCP calls. This is expected for the service, but it is still delegated account authority.

Skill content
mcporter config add kakiyo https://api.kakiyo.com/mcp --header "Authorization:Bearer USER_API_KEY"
Recommendation

Use a dedicated, revocable Kakiyo API key if possible, keep the local mcporter config protected, and revoke the key if the environment is shared or compromised.

What this means

A webhook pointed at the wrong or untrusted URL could expose prospect activity or campaign events.

Why it was flagged

The skill can configure webhook destinations for prospect-related events. This is a normal integration feature, but it can send lead or conversation-related events outside Kakiyo.

Skill content
**create_webhook** - Set up new webhook ... events:'["prospect.replied","prospect.qualified"]'
Recommendation

Only create or update webhooks after confirming the destination URL, event types, and expected payloads.

What this means

Resuming an agent may cause ongoing LinkedIn automation beyond the current chat turn.

Why it was flagged

The skill can control Kakiyo automation agents that may keep operating after the immediate request. This is disclosed and purpose-aligned, but users should treat it as persistent automation.

Skill content
**resume_agent** - Restart a paused agent
Recommendation

Confirm which agent is being resumed, what limits apply, and how to pause it again before enabling ongoing automation.

What this means

Users must trust the Kakiyo MCP endpoint and the mcporter client to handle their commands and API key appropriately.

Why it was flagged

The skill relies on a remote MCP server for tool behavior rather than local code in the artifact set. This is consistent with a SaaS integration, but it means the supplied artifacts do not expose the server-side implementation.

Skill content
mcporter config add kakiyo https://api.kakiyo.com/mcp
Recommendation

Verify the endpoint and provider before configuration, and use trusted installations of mcporter.