Back to skill

Security audit

SAP FICO Senior Consultant

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed SAP FI/CO reference and prompt skill with no executable payload, though its automatic trigger and French-only behavior may surprise some users.

Install this only where a French SAP FI/CO assistant is intended. Review the auto-trigger keyword list before enabling it in shared bots, because generic terms such as assessment, distribution, settlement, or closing could activate the SAP persona in mixed conversations.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (12)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

Forcing French technical SAP terminology at the skill level without regard to the user's language can create unsafe misunderstandings, especially for configuration, troubleshooting, or migration guidance. In a specialist enterprise domain, language mismatch can degrade comprehension and cause users to misapply instructions in production contexts.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill is configured to auto-activate on a broad list of generic SAP terms, including common words and transaction/table names that may appear in unrelated or mixed-context conversations. This can cause unintended prompt injection of the SAP expert persona, leading to context hijacking, reduced user control, and potentially incorrect handling of queries that only incidentally mention SAP-related tokens.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The trigger description states that the skill activates automatically when a message contains an SAP keyword, but it does not define precedence, confidence thresholds, exclusions, or how conflicts with other skills are resolved. Ambiguous activation behavior increases the chance of accidental invocation and makes the system easier to steer through crafted keyword-bearing inputs.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

Requiring responses in French as a quality criterion reinforces a non-consensual language constraint and may prevent the model from communicating critical SAP steps in the language the operator actually understands. In enterprise support scenarios, this raises the risk of operator error, especially for transactional or customizing instructions.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The examples and description are entirely in French, which biases the skill toward responding in a fixed language without evidence of user preference or an explicit localization requirement. This can cause incorrect-language responses, reduce usability, and create compliance or operational issues in multilingual environments where users expect language negotiation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The manifest sets "language": "fr", which establishes a fixed language for the skill. This is reinforced by the response schema requiring "format_language": "fr", and there is no indication that users can choose another language or opt in to French output.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The response schema explicitly requires outputs in French via "format_language": "fr". This is a natural-language policy concern because it enforces a locale-specific response format without any documented user choice or opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

Line L012 specifies 'Langues : Français technique SAP' and instructs the assistant to use SAP terminology in French. This imposes a specific language policy on all responses, and the file does not provide any user opt-in or alternative language handling.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

Line L135 requires 'Français technique' as a conduct rule, making the language constraint operational rather than merely descriptive. Because no opt-in, fallback, or exception is provided, this is a natural-language policy violation under the language/locale rule.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
82% confidence
Finding

The title and entire document content are written in French, but there is no natural-language indication that the skill is intentionally limited to French-speaking users or a specific locale. Under the language/locale policy, forcing a specific language without opt-in or documented justification can be a policy concern.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
94% confidence
Finding

The title and content indicate the skill file is presented entirely in French, but there is no natural-language indication that this language choice is optional or intentionally limited to a French-speaking audience. Under the policy, 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
95% confidence
Finding

This markdown file uses French throughout headings and descriptions, which can amount to a language/locale constraint for users who do not read French. The policy allows such constraints when the skill offers a language choice or clearly documents a justified regional scope, neither of which appears here.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.