deep-research skill

PassAudited by ClawScan on May 4, 2026.

Overview

This appears to be a structured research workflow that openly runs a local helper script, uses web research tools, and saves reports, with no artifact-backed malicious behavior found.

This skill looks purpose-aligned for deep research. Before installing, review the full orchestrator script, make sure bash and jq or python3 are available, choose where reports should be saved, and avoid confidential research topics unless you trust the search and webpage-fetch tools the agent will use.

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

Installing and using the skill will cause the agent to run a local bash script to manage research steps.

Why it was flagged

The skill requires the agent to execute a local shell script. This is clearly disclosed and central to the orchestration purpose, but users should know the skill is not purely prompt text.

Skill content
Agent 第一步必须 `bash orchestrator.sh "主题"` 初始化会话,后续每步通过 `--complete sX --next sY` 推进
Recommendation

Review the full local script before use and ensure you are comfortable allowing the agent to run it.

What this means

Research topics, keywords, and fetched URLs may be sent to external search or webpage retrieval services.

Why it was flagged

The workflow delegates searching to whatever search tools are available in the agent environment. That is expected for deep research, but the exact provider/tool boundary is not specified.

Skill content
使用当前环境可用的搜索工具执行网络搜索
Recommendation

Avoid using highly confidential topics unless you trust the available search/fetch tools, and confirm which tools the agent will use.

What this means

Research state and reports may remain in the skill directory or configured reports directory after the task is complete.

Why it was flagged

The script stores session state and reports on disk. This persistence is purpose-aligned, but it may retain business research topics and generated findings.

Skill content
CHECKPOINT_DIR="${SKILL_DIR}/.research_state"
REPORTS_DIR="${DEEP_RESEARCH_REPORTS_DIR:-${SKILL_DIR}/reports}"
Recommendation

Use `DEEP_RESEARCH_REPORTS_DIR` to choose an appropriate storage location and delete stored reports/state when no longer needed.

What this means

The skill may fail or behave differently if the needed local binaries are missing, even though the registry metadata says no binaries are required.

Why it was flagged

The metadata does not declare runtime binaries, while README.md documents dependencies on bash and jq or python3. This is a declaration gap, not hidden behavior.

Skill content
Required binaries (all must exist): none
Recommendation

Before installing, confirm bash and either jq or python3 are available.