Back to skill

Security audit

办公室 Token 洞察与提效助手

Security checks for vulnerabilities and agentic risk

Overview

This skill is a local office-productivity helper with disclosed filesystem use and no evidence of network, credential, persistence, or destructive behavior.

Before installing, be comfortable with a local tool that reads files you point it at, writes reports and ledgers you choose, and can optionally import bounded local host usage traces. Review ledger changes before confirming writes; pin optional export dependencies if you need reproducible builds.

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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (37)

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · CHANGELOG.md (reported line 223)May include surrounding context.

md
- **[C2] 加固安全预检红线(CI 闸门此前名不副实)**:
  - R1 补「动态导入危险目标模块」的检测(旧正则仅认常见的 eval / exec / 系统调用三类,动态导入可等价绕过)。仅匹配危险目标模块名,避免误伤「打印到标准错误流」等良性用法。
  - R2 把 socket 相关检测从单一构造函数名扩展为整个 socket 模块前缀(覆盖建连与连接等入口),并补 urllib3 / pycurl / grpc / websocket / httplib2 等客户端(旧正则漏掉裸 socket 联网)。
  - R3 放宽密钥长度门槛(8→4 字符),并新增对「api key / access token / client secret / private key / token / pwd」等**敏感名紧跟赋值号**的写法检测(完整模式清单以 `scripts/security_preflight.py` 的 R3 规则为准)。
  - **文档去夸大**:模块 docstring 明确这是「轻量静态启发式,非安全保证」,可被字符串拆分/环境变量/编码绕过,仅拦低级手滑,不在承诺中称「零风险」。
- **[R1] 会议纪要 `_DECISION_RE` 移除过宽关键词「确认」**(极常见中文词,导致「请确认参会」等行被误判为核心结论)。
- **[R3/R4] 清理死代码**:删除 `render_ppt_outline` 从未使用的 `chunks` 变量;移除 `executor.py` 未使用的 `field` 导入。

Intent-Code Divergence

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The README makes a strong privacy/safety claim that the skill is local-only and does not connect to the network, but later sections describe online SkillHub/ClawHub search capabilities. This is a security-relevant documentation integrity issue because users may expose prompts, metadata, or local usage context under a false assumption of no external transmission.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The boundary statement says the skill does not connect to external services, yet the evolution log later adds networked search clients. Contradictory trust-boundary statements can mislead users and reviewers about data exfiltration risk and cause deployment into environments that require strict offline operation.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The compliance section claims '全部本地处理,无外联', but the README elsewhere documents online clients and an online mode. False or stale compliance assertions are dangerous because they can cause operators to approve or install the skill in regulated or privacy-sensitive environments based on incorrect security properties.

Content

No source excerpt is available for this finding.

Direct flow: pathlib.Path.read_text (file read) → pathlib.Path.write_text (file write)

High
Category
Data Flow
Confidence
80% confidence
Finding

Data flows directly from a source (env vars, files, network) to a sink (network output, exec, file write) without intermediate validation.

Content

Scanner excerpt · scripts/ledger_agent.py (reported line 206)May include surrounding context.

python
if backup and p.is_file():
        ts = datetime.now().strftime("%Y%m%dT%H%M%S%f")
        backup_path = Path(str(p) + f".{ts}.bak")
        backup_path.write_text(p.read_text(encoding="utf-8"), encoding="utf-8")

    tmp = Path(str(p) + ".tmp")
    tmp.write_text(json.dumps(ledger, ensure_ascii=False, indent=2), encoding="utf-8")

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The document title and all release notes are written exclusively in Chinese, with no indication that language is configurable or that the skill is intentionally limited to a Chinese-speaking region. For a user-facing skill artifact, this constitutes a natural-language locale restriction without opt-in or documented justification.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The quickstart advertises very broad natural-language triggers such as '帮我写一份周报' and '整理一下会议纪要', which are common everyday phrases that may be spoken in contexts other than intentional skill invocation. In a host that routes conversational input automatically, this can cause unintended execution of the skill, potentially processing sensitive user content or initiating ledger-related actions without sufficiently explicit user intent.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

README 说明用户只需说“帮我写一份周报”“整理一下会议纪要”“分析这个 CSV”即可触发,这些短语都是常见自然办公表达,并未在此处给出明确的宿主范围、前缀要求或反例约束。虽然文档提到天禧 / OpenClaw 宿主,但对何时算技能触发、何时只是普通对话请求仍不够具体,存在误调用风险。

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

该表将多种自然语言句子列为“直接说出即可触发本技能”的提示词,但没有说明这些是示例还是完整触发集合,也没有定义宿主如何区分普通问答与技能调用。缺少排除条件和上下文限制会让调用范围显得过宽。

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

This markdown file is entirely written in Chinese and presents all user-facing release information only in that language. Under the policy rule for natural-language violations, forcing a specific language without opt-in or an explicit justified locale constraint is a reportable issue.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The skill’s manifest description is entirely in Chinese, and the rest of the file consistently presents triggers, examples, and usage only in Chinese. For a general-purpose office productivity skill, this creates a language/locale constraint without user opt-in or an explicit statement that the skill is intended only for Chinese-speaking users.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

The document title and all operational guidance are presented exclusively in Chinese, which effectively imposes a specific language for this skill artifact. Under the stated policy, forcing a language or locale without user opt-in can be a natural-language policy violation unless clearly documented as 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
96% confidence
Finding

This code embeds all user-facing documentation and examples in Chinese, and later user-facing prompts and command examples are also Chinese-only. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale constraint is clearly documented and justified.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The handler and execution flow return Chinese-only status messages, confirmations, warnings, and examples, which effectively require users to interact in Chinese. The file does not provide a user choice of language or a documented justification for restricting interaction to this locale.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

This code file contains primary documentation and multiple user-facing messages exclusively in Chinese, which can effectively force a specific language on downstream users or operators. The policy allows locale constraints only when user choice is offered or the constraint is clearly documented and justified, neither of which appears here.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The documentation states a strict write boundary limited to the user-specified output path and says the skill does not scan files outside the skill directory. In practice, the code reads arbitrary filesystem paths from --input and CSV text/path inputs, and writes to arbitrary paths through --init-ledger, --apply-ledger, and output export helpers, so the stated red line is contradicted.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The function docstring and user-facing fallback/error text explicitly state and implement Chinese-only output, and the rest of the file consistently returns Chinese responses. This creates a language policy concern because the skill forces a specific language without user opt-in or a documented justification that it is region-specific.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The module docstring and embedded section/label requirements are entirely in Chinese and encode Chinese-specific headings such as “本周概览”, “风险与阻塞”, and other task labels. There is no indication that the skill offers a language choice or that the Chinese-only constraint is justified as a region-specific tool, which creates a natural-language locale policy concern.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

This JSON manifest defines activation-style keyword mappings using very generic terms such as "文档", "文章", "数据", "代码", "开发", "演示", and "邮件". These words are common in everyday conversation and the file provides no explicit constraints or negative examples, increasing the risk of unintended matches or incorrect skill/type selection.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

This markdown file is entirely written in Chinese, including the invocation examples, explanations, and output demonstrations, and it does not indicate that the skill is region-specific or that users may choose another language. Under the language/locale policy, forcing a specific language without opt-in is a natural-language policy concern.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
89% confidence
Finding

At L030 the text explicitly states there were no code behavior changes, yet L038 describes adding a new test class to enforce document-pattern checks. Tests are code and do change repository behavior in CI/verification terms, so the documentation overstates the absence of code changes.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

L049 says the release was purely documentation/packaging with zero code behavior change, but L054 states a test file assertion was changed. That is not a runtime capability change, but it does contradict the blanket statement that only docs/packaging changed.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
63% confidence
Finding

L062 says all external behavior remained unchanged, yet L068 describes deleting a CRITICAL rule and related fields from recommendation logic. Although the text argues the rule was unreachable, removing recommendation rules is a semantic code change and makes the 'all external behavior unchanged' claim too absolute.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
79% confidence
Finding

The JSON values for task type and notes are entirely in Chinese (for example at L05, L10, L14, L19, L23, and L28), which can indicate a language-specific skill configuration without any visible opt-in or documented locale justification in this file. Under the policy, forcing a specific language without user choice can be a natural-language policy concern.

Content

No source excerpt is available for this finding.

Unpinned Dependencies

Low
Category
Supply Chain
Confidence
95% confidence
Finding

The optional dependency is specified with only a lower bound, which allows future installs to resolve to different versions over time. This weakens supply-chain reproducibility and makes it harder to ensure that a known-safe version of python-docx is installed, especially given historical XXE-related advisories in older releases.

Content

Scanner excerpt · requirements-optional.txt (reported line 4)May include surrounding context.

text
# 可选依赖(不属于技能核心零依赖默认)
# 仅在使用 `scripts/executor.py --format docx/xlsx` 导出 Word / Excel 时需要。
# 未安装时 executor 自动降级:docx → Markdown,xlsx → CSV,功能不中断。
python-docx>=1.0
openpyxl>=3.1