Deep Research Strategy
PassAudited by ClawScan on May 10, 2026.
Overview
This instruction-only research skill appears purpose-aligned, but users should know it is designed to delegate queries to multiple search agents or external search tools.
This looks like a benign instruction-only deep research coordinator. Use it when you want multi-source research, but avoid giving it confidential or personal topics unless you are comfortable with those topics being sent to available search agents or external search providers.
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 may be less likely to challenge or reinterpret the user’s stated research terms, even when they seem unfamiliar.
The skill strongly prioritizes user-provided terms and instructions. In context, this appears intended to prevent the agent from incorrectly changing research topics, but the wording is broad enough that users should understand it changes how the agent treats user input during research.
Golden Rule 1: User Input is the Absolute Source of Truth ... This is your highest command and must be obeyed unconditionally.
Keep the rule scoped to preserving research terms and require normal safety, verification, and source-quality checks to remain in force.
A simple research request may be sent to several search agents or tools instead of being answered directly.
The skill mandates multi-agent/tool delegation for every task. This is coherent with a deep research strategy skill, but it may be more tool use than some users expect for simple queries.
Any user task, no matter how simple it seems, must be decomposed into at least more than 2 and less than 6 parallel sub-tasks and assigned to different subordinate agents.
Use this skill for tasks where external research and multi-source delegation are expected; avoid invoking it for private or simple questions that do not need external search.
Research topics may be shared with separate search agents or external search systems if those tools are available.
The skill describes delegating user research topics to several subordinate agents and search tools. That inter-agent flow is central to the skill, but the provided artifacts do not define data boundaries or identities for those agents.
assign them to your subordinate expert Agents ... Internet Searcher ... Twitter Information Searcher ... Financial Information Searcher ... Academic Research Searcher
Do not submit sensitive, confidential, or personal research topics unless you are comfortable with them being delegated to search tools.
The review cannot confirm how the referenced search agents behave, what providers they call, or how they handle submitted queries.
The instruction file references external or subordinate tools that are not present as code files or declared capabilities in the provided artifacts. This is not suspicious by itself for an instruction-only skill, but it limits what can be verified from the supplied package.
Corresponding Tool: `call_web_search_agent` ... `Tech_and_Internet_Domain_Search_Agent` ... `Finance_Search_Agent` ... `call_academic_search_agent`
Before using the skill, confirm which search tools are available in your environment and whether they meet your privacy and provenance expectations.