v1.0.2

Technical Insight

ReviewClawScan verdict for this skill. Analyzed May 1, 2026, 8:29 AM.

Analysis

The skill mostly matches its technical-analysis purpose, but it needs review because it can run local helper commands, references an API key, and contains packaging/path inconsistencies that are not fully disclosed in the registry metadata.

GuidanceBefore installing or running this skill, confirm that you are comfortable with local command execution for diagram generation, review or supply the missing drawio helper explicitly, declare and scope any Tavily API key, and avoid using it on proprietary repositories unless you are comfortable with generated analysis files being stored under the OpenClaw workspace.

Findings (6)

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.

Abnormal behavior control

Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.

Agentic Supply Chain Vulnerabilities

SeverityMediumConfidenceHighStatusConcern

architecture-workflow-v2.py

subprocess.run(['python3', "scripts/drawio-generator.py", temp_analysis_path, self.diagrams_dir], capture_output=True, text=True, check=True)

The workflow executes a helper script under scripts/drawio-generator.py, but that helper is not listed in the provided file manifest; this creates a provenance gap for code that may be run during diagram generation.

User impactIf this workflow is used, diagram generation may fail or may depend on code that was not included and reviewed with the skill package.

RecommendationInclude the referenced helper in the package, declare it in metadata/install requirements, and ensure users approve local helper execution.

Unexpected Code Execution

SeverityLowConfidenceHighStatusNote

architecture-diagram-generator.py

subprocess.run(['plantuml', '-tpng', diagram_path], check=True, capture_output=True)

The diagram generator invokes a local PlantUML command. This is aligned with architecture diagram generation, but it is still local command execution and the registry does not declare PlantUML as a required binary.

User impactUsing the diagram feature may run local tooling on generated diagram files.

RecommendationVerify that PlantUML is installed from a trusted source and that command execution is expected before running diagram generation.

Human-Agent Trust Exploitation

SeverityLowConfidenceHighStatusConcern

SKILL-FOUR-PLUS-ONE.md

export WORKSPACE_DIR="/home/Vincent/.openclaw/workspace"

A bundled automatic workflow still contains a developer-specific absolute path, which conflicts with README/CLAWHUB_READY statements that /home/Vincent references were removed.

User impactUsers may trust the stated path-safety fix even though a packaged workflow can still point at a developer-specific location.

RecommendationRemove or update stale workflow documentation, replace hardcoded home paths with user-relative paths, and avoid safety claims until the whole package is consistent.

Permission boundary

Checks whether tool use, credentials, dependencies, identity, account access, or inter-agent boundaries are broader than the stated purpose.

Identity and Privilege Abuse

SeverityLowConfidenceHighStatusNote

DEVELOPMENT_PLAN.md

配置文件: `~/.config/tech-insight/tavily_key`

The artifacts mention a local Tavily API-key file, while the registry requirements declare no primary credential or required environment variables.

User impactThe skill may require or use an API key even though the install metadata does not make that clear.

RecommendationDeclare the Tavily credential in metadata, document what it is used for, and use a narrowly scoped key if possible.

Sensitive data protection

Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.

Memory and Context Poisoning

SeverityLowConfidenceHighStatusNote

SKILL.md

所有深度分析结果自动保存到：`~/.openclaw/workspace/tech-insight/technical-insight/{技术名称}/` ... `code-analysis.json`

The skill intentionally persists generated reports, diagrams, sources, and code-analysis data under the local OpenClaw workspace.

User impactIf used on proprietary or sensitive repositories, technical details may remain on disk after the task finishes.

RecommendationUse the skill only on code and architecture information you are comfortable storing locally, and delete the generated workspace folder when no longer needed.

Insecure Inter-Agent Communication

SeverityLowConfidenceHighStatusNote

SKILL.md

**调用 source-to-architecture 技能**: 使用优化后的源码到架构技能进行专业架构图生成

The main skill instructions say it calls another skill for architecture generation; that is purpose-aligned, but metadata does not describe the downstream skill boundary or what data is handed off.

User impactRepository or analysis details may be passed to another skill as part of normal operation.

RecommendationDocument downstream skill use and ask for user confirmation before passing sensitive project details to other skills.