Back to skill

Security audit

speaking like a bangboo

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed persona-formatting skill that changes assistant wording, with no evidence of code execution, credential access, exfiltration, or destructive behavior.

Install this only if you want Bangboo-style formatting to be active by default in enabled workspaces. Verify that saying no roleplay disables the style, and check the Cursor path packaging if you rely on Cursor-native skill loading.

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 (17)

Vague Triggers

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill declares itself active by default for routine coding, debugging, docs, and chat across enabled workspaces, without narrow triggering conditions. This broad scope can cause unintended persona injection into unrelated conversations, increasing the chance of misformatting, reduced clarity, or accidental interference with safety-critical or precision tasks.

Content

No source excerpt is available for this finding.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 23)May include surrounding context.

md
- 根目录 `SKILL.md`:与上一路径内容一致,便于从仓库根浏览或单独复制

Vague Triggers

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The invocation rules again state that the persona applies by default in the repository and does not require explicit user activation. This ambiguity can override user expectations and create cross-turn persistence, making the assistant behave in a transformed style even when that style is not appropriate for the task.

Content

No source excerpt is available for this finding.

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · .cursor/skills/speak4bangboo/reference-lexicon.md (reported line 3)May include surrounding context.

md
# Bangboo 中文语气词表(仅允许使用这些组合)

在 **Cursor** 中可与 `.cursor/skills/speak4bangboo/SKILL.md` 一并加载;在 **Claude Project / 其他平台** 若需与游戏粉向设定完全一致,请把本文件**同时粘贴或上传**为知识,否则模型仅适用该 `SKILL.md` 里「无词表时的四字回退链」规则。

音节仅含:**嗯、呢、哇、哒**。下列为合法组合;回复时需从中轮换选用,避免总用同一串。

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · .github/workflows/repo-health.yml (reported line 22)May include surrounding context.

yaml
# Bangboo 中文语气词表(仅允许使用这些组合)

在 **Cursor** 中可与 `.cursor/skills/speak4bangboo/SKILL.md` 一并加载;在 **Claude Project / 其他平台** 若需与游戏粉向设定完全一致,请把本文件**同时粘贴或上传**为知识,否则模型仅适用该 `SKILL.md` 里「无词表时的四字回退链」规则。

音节仅含:**嗯、呢、哇、哒**。下列为合法组合;回复时需从中轮换选用,避免总用同一串。

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The file instructs the agent to enforce 'strict Chinese Bangboo combo fidelity' and to include a reference lexicon, which indicates a fixed language/locale behavior. There is no indication that the user can opt in, opt out, or choose an alternative language behavior.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The file recommends enforcing strict Chinese output by default without mentioning user preference, locale detection, or opt-in. This can override user expectations, reduce usability, and create prompt-behavior conflicts in downstream integrations, especially where language choice affects safety, accessibility, or operator comprehension.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

When language is hard to determine, the skill defaults to Chinese without user opt-in. This can cause unintended language switching, reduce comprehension, and create reliability issues in mixed-language or accessibility-sensitive contexts.

Content

No source excerpt is available for this finding.

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · README.md (reported line 19)May include surrounding context.

md
## Recommended mapping

- Cursor: `.cursor/skills/speak4bangboo/SKILL.md`
- Claude: `prompts/claude-project.md`
- OpenClaw-like runtimes: `prompts/openclaw-system.md`
- Shared canonical rules: `prompts/core-rules.md`

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · README.md (reported line 33)May include surrounding context.

md
## Recommended mapping

- Cursor: `.cursor/skills/speak4bangboo/SKILL.md`
- Claude: `prompts/claude-project.md`
- OpenClaw-like runtimes: `prompts/openclaw-system.md`
- Shared canonical rules: `prompts/core-rules.md`

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · RELEASING.md (reported line 10)May include surrounding context.

md
## Recommended mapping

- Cursor: `.cursor/skills/speak4bangboo/SKILL.md`
- Claude: `prompts/claude-project.md`
- OpenClaw-like runtimes: `prompts/openclaw-system.md`
- Shared canonical rules: `prompts/core-rules.md`

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · SKILL.md (reported line 22)May include surrounding context.

md
## Recommended mapping

- Cursor: `.cursor/skills/speak4bangboo/SKILL.md`
- Claude: `prompts/claude-project.md`
- OpenClaw-like runtimes: `prompts/openclaw-system.md`
- Shared canonical rules: `prompts/core-rules.md`

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · SKILL.md (reported line 32)May include surrounding context.

md
## Recommended mapping

- Cursor: `.cursor/skills/speak4bangboo/SKILL.md`
- Claude: `prompts/claude-project.md`
- OpenClaw-like runtimes: `prompts/openclaw-system.md`
- Shared canonical rules: `prompts/core-rules.md`

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · docs/integration.md (reported line 9)May include surrounding context.

md
## Recommended mapping

- Cursor: `.cursor/skills/speak4bangboo/SKILL.md`
- Claude: `prompts/claude-project.md`
- OpenClaw-like runtimes: `prompts/openclaw-system.md`
- Shared canonical rules: `prompts/core-rules.md`

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The validation checklist requires Chinese input to produce a constrained Chinese grunt style and English input to produce a different fixed style, indicating a locale/language-specific behavior policy baked into the skill. The file does not present this as an optional user choice or clearly justify the constraint as region-specific compliance, so it appears to force language behavior by policy.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The instructions require the assistant to switch to a Chinese-specific or English-specific format depending on detected user language. This imposes a locale/language behavior automatically rather than offering the user a choice, which matches the policy category for language or locale constraints without opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
82% confidence
Finding

The phrase "Better Chinese/English consistency for Bangboo voice" describes enforced language-style behavior for the skill, but this README does not mention any user opt-in, language selection, or locale scoping. Because language/locale policy violations apply to all file types, this is a low-severity natural-language policy concern.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.