Deep Strategy
PassAudited by ClawScan on May 1, 2026.
Overview
This instruction-only planning skill is mostly coherent, but users should notice that it strongly prioritizes user wording and encourages delegation of tasks and attachments to named sub-agents.
This skill appears safe to install as an instruction-only planning aid. Use caution with sensitive attachments, and consider asking it to confirm before delegating files or using research/editor sub-agents.
Findings (3)
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The agent may preserve the user's exact terms or assumptions even when they may be outdated or incorrect.
The skill tells the agent to heavily prioritize user-provided wording and facts. This appears intended to preserve user intent, but the absolutist framing could over-constrain independent validation.
Your ultimate mission is **to complete tasks ... under the premise of absolute loyalty to user input**.# Golden Rule 1: User Input is the Absolute First Fact**This is your highest, inviolable instruction.**
Use clear instructions and ask the agent to verify facts when accuracy matters; skill authors should clarify that fidelity to user input does not override safety or higher-priority instructions.
Tasks may involve extra tool or sub-agent calls, which could affect cost, latency, or how information is processed.
The skill mandates specific tool/sub-agent delegation for broad task types. Delegation is central to the stated purpose, but users should know the skill may route work through other tools rather than handling it directly.
To complete a proposal/strategy writing task, you must first call `deep_research`, and then call `chief_editor`... All data analysis processing, table parsing processing, and code tasks must be given to this subordinate
Before using it for sensitive or costly work, confirm which delegated tools are available and whether the agent should ask before invoking them.
If you provide sensitive documents, they may be shared with the delegated assistant tools needed to complete the task.
The instruction can cause user-provided files to be forwarded to delegated tools/sub-agents. This is purpose-aligned, but the artifact does not define detailed data-boundary or approval rules for attachments.
When delegating tasks to subordinates, be sure to pass all required attachments to them in `attached_files`.
Avoid attaching sensitive files unless necessary, and ask the agent to confirm before forwarding attachments to sub-agents.