Decomposes complex user requests into executable subtasks, identifies required capabilities, searches for existing skills at skills.sh, and creates new skills when no solution exists. This skill should be used when the user submits a complex multi-step request, wants to automate workflows, or needs help breaking down large tasks into manageable pieces.
ReviewAudited by ClawScan on May 10, 2026.
Overview
This planning skill is coherent, but it may steer the agent to globally install or create other skills and set scheduled automations without clear approval safeguards.
Use this skill only if you want the agent to plan across other skills. Before allowing it to install or create anything, ask to review each proposed skill, avoid global auto-confirm installs, verify the source, and set clear limits for any scheduled automation.
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.
The agent could add new capabilities or instructions that affect future sessions and tasks, potentially beyond what the user intended.
The documented workflow can globally install discovered skills while suppressing confirmation prompts, but the artifacts do not clearly require user approval before this high-impact environment change.
# Install a discovered skill npx skills add <owner/repo@skill> -g -y
Require explicit user confirmation before installing or creating skills, avoid auto-confirm flags by default, and prefer scoped or temporary installs when possible.
A low-quality or malicious third-party skill could be introduced into the agent environment if search results are installed without review.
The skill is designed to discover and introduce other skills, but the artifacts only mention general quality checks and do not define provenance, pinning, security review, or trust boundaries for external skills.
- **Skill Search**: Searches the [skills.sh](https://skills.sh/) ecosystem for existing solutions - **Skill Creation**: Generates new skills when no existing solution is found
Inspect any discovered skill before installation, prefer trusted publishers and pinned versions, and include security/provenance checks in the workflow.
Automations may keep running and interacting with services unless the user sets clear limits or cleanup steps.
The skill may plan persistent scheduled automation. This is aligned with workflow automation, but scheduled jobs can continue operating after the immediate task.
name: "Configure scheduled execution" ... output: "Active scheduled job"
Before enabling scheduled workflows, confirm the schedule, scope, stop condition, and removal procedure.
Using broad credentials could expose email, messaging, or workspace access to newly installed or generated skills.
Example workflows may require account credentials or service tokens. This is expected for email and Slack automation, but those credentials can grant meaningful account access.
input: "Email credentials/session" ... input: "Summary text, Slack webhook/token"
Use least-privilege tokens, avoid sharing long-lived session credentials, and review which skill will handle each credential.