Back to skill

Security audit

Agent Migration Pack Openclaw

Security checks for vulnerabilities and agentic risk

Overview

This skill is a legitimate migration toolkit, but it packages and encourages long-term storage of sensitive personal and third-party data with weak controls.

Review before installing or sharing any generated pack. Treat generated folders and ZIP files as sensitive backups: remove bundled real examples, redact personal and third-party details, avoid sharing relationship/contact data without consent, and verify whether skill inventories are being included even when you did not request them.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (2)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/migrate.py:127
Finding

Sensitive Personal and Third-Party Data Is Included in Migration Archives by Default

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Note
Location
scripts/generate-pack.py:143
Finding

The --include-skills Privacy Control Is Ignored

Content
View full analysis
dict: skills_dir = self.workspace_root / "skills" template_dir = self.workspace_root / ".skills" skills = { "template_version": self.VERSION, "file_type": "skills", "catalog_version": "1.0", "export_date": self.timestamp, "extracted_at": self.timestamp, "skills": [], "recommended_skills": [] } for skills_path in [skills_dir, template_dir]: if skills_path.exists() and skills_path.is_dir(): skills["source_files"] = skills.get("source_files", []) skills["source_files"].append(str(skills_path)) for item in skills_path.iterdir(): if item.is_dir(): skills["skills"].append({ "name": item.name, "path": str(item), "status": "installed" }) ``` The generation workflow invokes extraction and writes the catalog regardless of the `include_skills` argument: ```python def generate_pack(self, output_dir: str, include_skills: bool = True) -> dict: identity = self.extract_identity() owner = self.extract_owner() memory = self.extract_memory() skills = self.extract_skills() files_to_save = [ ("IDENTITY/identity.json", identity), ("OWNER/owner.json", owner), ("MEMORY/core-memory.json", memory), ("SKILLS/catalog.json", skills), ] ``` However, the command-line interface states that skills are excluded by default: ```python parser.add_argument( " ...[truncated 2167 chars]
Remediation
View remediation
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (28)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The README explicitly states that the example contains real personal information and is only for reference, but it does not clearly instruct users not to commit, share, publish, or redistribute those files. In a migration-package example containing identity, owner, memory, and relationship data, that omission materially increases the risk of accidental privacy leakage and downstream exposure of sensitive personal data.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

This manifest includes user-facing source and description fields in Chinese, but provides no indication that the skill catalog is intentionally region-specific or that users can opt into another language. Under the policy, forcing a specific language without opt-in can be a natural-language policy violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

This JSON file’s user-facing natural-language content is entirely in Chinese, with no indication that the skill is region-specific or that users can opt into another language. Under the policy, forcing a specific language without user choice is a natural-language policy violation unless the locale constraint is explicitly justified.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The natural-language content throughout the memory file is written entirely in Chinese, including the description, team mode, business context, schedules, and judgments. There is no indication that this language choice is optional, user-selected, or justified as a region-specific requirement, which can violate language/locale policy expectations.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The conversation tone is explicitly specified in Chinese ("口语化,微信聊天风格"), which imposes a language/locale preference in the skill configuration. Under the policy, forcing a specific language without offering the user a choice or documenting an opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The guide explicitly instructs users to collect and persist extensive private user data and third-party personal data from conversation history into migration files. Even though it warns against storing passwords and suggests redaction, it still normalizes broad data aggregation and transfer, increasing the risk of privacy violations, over-collection, unauthorized disclosure, and downstream compromise if the package is mishandled.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The memory template directs extraction of sensitive information from emails, task records, dialogue history, and platform records into long-term structured memory. This creates a durable concentration of sensitive data that may exceed original context and retention expectations, making later leakage, misuse, or unauthorized cross-context access more damaging.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The relations section tells operators to extract and migrate third-party contact details, trust levels, communication summaries, and project history from platform conversation records. This is particularly risky because it involves data about other people who may not have consented, enabling privacy harm, profiling, and disclosure of relationship metadata if the migration package is shared or breached.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
89% confidence
Finding

The skill describes local file copy, validation, packaging, and checksum workflows, which imply file read/write capability, but it declares no explicit tool scope or permissions boundary. In an agent ecosystem, undeclared file access can cause users or host platforms to grant broader access than intended and makes security review and sandboxing harder.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

This skill encourages collecting and packaging highly sensitive data such as owner details, memories, relationships, and installed skills for backup or sharing, but provides no privacy warning, minimization guidance, or handling safeguards. That creates a real risk of over-collection, accidental disclosure, and unsafe sharing of personal or confidential information.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The template description and examples are written entirely in Chinese, including operational guidance such as time estimate and field descriptions, with no indication that language choice is optional. This can represent a language policy issue because the skill content appears to assume a fixed locale rather than offering user choice or documenting a justified region-specific constraint.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

All user-facing descriptions and instructions in the template are written only in Chinese, with no indication that the user can choose another language or locale. This is a natural-language policy issue because it effectively forces one language without documenting user opt-in or a justified region-specific constraint.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The description and instructional content are written entirely in Chinese, and the template gives no indication that another language is allowed or that the Chinese-only locale is required for a specific regional or compliance purpose. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

This JSON template includes user-facing descriptions and instructions entirely in Chinese, such as the file description and comment instructions. Because the file does not offer an opt-in language choice or explain that it is intended only for a Chinese-language environment, it can violate the language/locale policy for broadly distributed skills.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The entire style guide is written as a communication standard in Chinese and repeatedly defines response structure, tone, and examples only in Chinese, indicating a forced language/locale preference. The file does not offer user opt-in, alternative language choices, or state that this is a region-specific tool with a justified locale constraint.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The script copies data derived from USER.md and MEMORY.md into a new migration package, which can duplicate sensitive personal or operational information into a broader, easier-to-share artifact. Because this happens automatically and the skills export behavior is also misleading elsewhere, users may unintentionally package confidential content and later leak it through backups, uploads, or transfers.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The argument help text states --include-skills means '包含技能目录 (默认: 不包含)', implying skills should be omitted unless the flag is set. However, generate_pack unconditionally calls extract_skills() and always writes SKILLS/catalog.json, and the include_skills parameter is never used, so the documented behavior contradicts the actual behavior.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
72% confidence
Finding

Natural-language policy checks apply to all file types. The document presents all instructions and safety guidance only in Chinese, and it does not state that the skill is region-specific or provide an alternative language option, which can be interpreted as forcing a specific language without opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
80% confidence
Finding

The file encodes all personal profile and preference content in Chinese, including schedule and preference fields, which implies a fixed language/locale context. Because there is no accompanying indication that the user opted into Chinese-only handling or that the skill is explicitly region-scoped, this can violate the policy against forcing a specific language without user opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

This markdown file contains user-facing instructional content exclusively in Chinese, and it does not indicate that other languages are supported or that Chinese is optional. Under the policy rule for natural-language violations, forcing a specific language without user opt-in can be a locale-policy issue.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The example owner.json sets the language preference to 'zh-CN' as a default example without indicating that language should be chosen by the user. This can be interpreted as forcing a specific locale rather than offering an explicit language choice.

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
83% confidence
Finding

This JSON manifest/template includes a very broad role description such as 'AI助手、数字员工' responsible for search, product analysis, collaboration, and social management. In a manifest-like file, such expansive capability wording lacks clear scope boundaries or exclusion conditions, which can contribute to overly broad skill matching or invocation intent.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

This JSON template uses Chinese throughout its description, examples, and instructions, which effectively imposes a specific language on users. The file does not provide an opt-in language choice or explain that the skill is intentionally limited to a Chinese-language context.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
86% confidence
Finding

This JSON manifest contains user-facing descriptive text in Chinese, including the top-level description and time estimate, and that pattern continues throughout the file. Because the file provides no opt-in, alternative locale, or justification for being China-specific, it appears to impose a specific language/locale by default.

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
89% confidence
Finding

The template instructs population from broad sources including conversation history, which can cause over-collection of personal data without clear scope, consent boundaries, or minimization rules. In a private owner profile template containing identity, family, business, and preference fields, this ambiguity can lead to unintended ingestion of sensitive information and privacy violations.

Content

No source excerpt is available for this finding.