Back to skill

Security audit

Bookkeeping Reconciliation Helper

Security checks for vulnerabilities and agentic risk

Overview

This is a documentation-only bookkeeping reconciliation helper with broad triggers but no hidden execution, data access, persistence, or destructive behavior.

Install this if you want an agent to help structure bookkeeping reconciliation work. Because it can be invoked implicitly from broad bookkeeping terms, use it intentionally for reconciliation tasks and avoid sharing unnecessary bank, invoice, receipt, or ledger details unless they are needed for the task.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (11)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger phrases are broad enough to activate on generic finance- or workflow-related requests, which can cause the skill to run outside its intended scope. In an agent system, ambiguous activation boundaries increase the chance of unintended instruction injection into unrelated conversations, leading to user confusion, poor task routing, or unsafe automation decisions.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
83% confidence
Finding

This file presents the user-facing README in Chinese, while key operational content such as the trigger phrases and workflow description remain mixed-language, and there is no statement that language selection is optional or region-specific. Under the policy, forcing a specific language or locale without opt-in can be a natural-language policy violation.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The trigger phrases are broad, repetitive, and loosely scoped, which can cause the skill to activate for requests that only partially match bookkeeping or reconciliation intent. In an agent environment, overbroad invocation increases the chance of unintended context capture, misrouting, or execution of workflows on unrelated financial data, creating avoidable security and privacy risk.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest description defines activation conditions with a long list of broad domains and output types such as 'software-and-data', 'bookkeeping', 'analysis', and 'implementation support', which are ambiguous and likely to match many unrelated requests. This can lead to unintended skill selection, reducing routing precision and potentially causing the agent to follow the wrong operational guidance for sensitive finance-related tasks.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The trigger examples are written in broad, natural-language phrases that overlap with ordinary bookkeeping requests, which increases the chance the skill is invoked when the user did not explicitly intend it. In an agent system, overbroad activation can cause prompt-routing errors, expose irrelevant internal workflow instructions, or override a more appropriate skill for the task.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill description is very broad and can activate for common finance-related requests without clear boundaries on when it should or should not be used. Overbroad activation increases the chance of unintended invocation, which can override more appropriate skills or cause the agent to apply bookkeeping workflows to unrelated user intents.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger keywords include generic terms like 'software-and-data', 'bookkeeping', and 'receipts', which are likely to appear in many ordinary requests outside the intended reconciliation workflow. This creates a collision risk where the skill may be selected too often, leading to misrouting, irrelevant guidance, or accidental exposure of financial-processing logic in the wrong context.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The default prompt uses very broad natural-language trigger wording tied to common finance/bookkeeping terms, which can cause the skill to activate in routine conversations where the user did not explicitly intend to invoke it. This increases the risk of over-collection of sensitive financial context, unintended workflow steering, or accidental use in situations involving invoices, receipts, bank data, and ledgers.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger sentences are so generic and malformed that they could match ordinary user requests unrelated to this specific bookkeeping workflow, causing the skill to activate unexpectedly. Over-broad activation increases the chance of prompt routing errors, irrelevant guidance, or interference with other more appropriate skills, especially in a system that auto-selects skills from natural-language triggers.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
71% confidence
Finding

SQP-3 applies to all file types and includes language/locale policy violations. The file lists separate English and Chinese instruction and guide files, but does not state that the user may choose their preferred language or that language selection is opt-in, which can be interpreted as a locale handling policy gap.

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The example trigger phrases are ambiguous and only show positive examples, without distinguishing valid invocations from ordinary finance questions that should not activate the skill. This ambiguity can train selection systems or users toward over-invocation, increasing false positives and reducing routing accuracy.

Content

No source excerpt is available for this finding.