Back to skill

Security audit

法务助手(Legal Assistant)

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed Chinese-language legal workflow skill that handles contracts, drafts analysis, and records ledger/archive outputs with human review controls, with no evidence of hidden code execution or deceptive behavior.

Install only in an environment where Feishu access, Hermes Agent permissions, audit logs, archives, and contract ledger writes are scoped to authorized legal/business users. Before production use, confirm that ledger updates and reminders are drafts or require explicit approval, and define retention/redaction rules for sensitive contract data.

Vulnerability Patterns
  • 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
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Rogue AgentSelf-Modification, Session Persistence
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (18)

Self-Modification

High
Category
Rogue Agent
Confidence
85% confidence
Finding

Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.

Content

Scanner excerpt · employee.yaml (reported line 107)May include surrounding context.

yaml
- L05_revision_suggestion
- L06_legal_qa
- L07_approval_opinion
- L08_contract_ledger_update
skill_catalog: skills/legal_skills.yaml
approval_policy:
  requires_human_confirmation:

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill advertises automatic contract-ledger updates and reminder-task generation after pre-review or workflow progression, but it does not require explicit user confirmation before modifying records or creating tasks. In a legal workflow, silent state changes can create inaccurate compliance records, trigger unauthorized notifications, or cause users to believe an action was officially recorded when it was only suggested by the model.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The manifest presents the skill name, owner, deployment channel, and mission entirely in Chinese, indicating the skill is designed to operate in a single language. There is no visible statement offering users a language choice or documenting a justified region-specific constraint, which can violate language/locale policy requirements.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
84% confidence
Finding

The manifest presents all user-facing skill names and purpose descriptions in Chinese only, with no indication that the user can choose another language or that the skill is intentionally limited to a Chinese-language legal workflow. This creates a natural-language locale constraint that is undocumented and not clearly justified in the file.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The prompt is written entirely in Chinese and defines the assistant role and behavior only in Chinese, implying a fixed language/locale for operation. The file does not mention that users may choose another language or that the Chinese-only constraint is optional or region-justified.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

This markdown template is entirely written in Chinese, including the report title and all section headings, which indicates the skill may require output in a specific language. The provided file does not offer a language choice or explain a justified region-specific constraint, so it appears to violate the language/locale policy requirement.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The template content is entirely in Chinese, including the title and table headings, with no indication that language selection is optional or that the skill is intended only for a Chinese-language or region-specific workflow. This can violate a language/locale policy when users are not given a choice or opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

The file is entirely written in Chinese and presents the interaction pattern and expected outputs only in that language, without offering any user language choice or noting that the skill is region- or locale-specific. Under the policy criteria, this can be a language/locale constraint that is imposed implicitly rather than explicitly opted into.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The workflow explicitly accepts user-submitted text and attachments, parses them, retrieves related materials, and then logs and archives the results, but it does not state any user-facing warning, consent step, data minimization rule, or restriction for sensitive legal content. Because contracts commonly contain personal data, financial terms, trade secrets, and privileged information, silent persistence and broad processing increase the risk of unintended disclosure, over-collection, or retention of sensitive data.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The contract ledger workflow states that contract status, amount, term, risk labels, and reminders will be maintained and updated, yet it does not clearly warn users that submitted contract data may be persisted in a ledger system. In a legal-assistant context, this is particularly sensitive because ledger entries can centralize commercially sensitive and personal information, making accidental retention, unauthorized access, or secondary use more likely.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
82% confidence
Finding

The README title and all operational descriptions are written only in Chinese, with no indication that users may choose another language or locale. Under the stated policy, forcing a specific language without opt-in can be a natural-language policy violation unless the locale constraint is documented and justified.

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, including markdown. The document consistently presents the assistant interaction and workflow only in Chinese, with no indication that language is configurable or that this locale restriction is intentional and documented.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
83% confidence
Finding

This is a markdown file, so missing user warnings apply to described behaviors that could affect user data or system integrity. The step says results are archived and a record is returned, but the document does not warn that task outputs and related materials may be stored or retained.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

The README presents all user-facing instructions exclusively in Chinese, and there is no indication that the skill is limited to a Chinese-speaking audience or that users can choose another language. This can violate a language/locale policy when skills are expected to avoid 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
92% confidence
Finding

The file title and all operational instructions are written exclusively in Chinese, which imposes a specific language on operators. The runbook does not state that the skill is intentionally limited to Chinese-speaking users or offer any language/locale choice, which matches the policy-violation category for forced language without opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

The README content, including the title and all descriptions, is presented exclusively in Chinese. Under the stated policy, forcing a specific language without user opt-in can be a natural-language policy violation unless the locale restriction is explicitly justified.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

The heading "法务助手归档记录" presents the template in Chinese with no indication that language selection is optional or limited to a Chinese-specific context. This can violate language/locale policy when a skill implicitly forces one language for all users.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

This markdown template uses Chinese for all labels and headings, which can impose a specific language on users without any stated option to choose another locale. The policy allows language constraints only when users are given a choice or the locale restriction is clearly documented and justified.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.