Back to skill

Security audit

专利合作关系

Security checks for vulnerabilities and agentic risk

Overview

The patent workflow fits its purpose, but weak credential and file-path safeguards warrant review.

Queries go to PatentMax and may incur disclosed charges. Keep confidential invention details out of searches and avoid sharing paid keys through chat. Before supplying a paid key, request safer destination and redirect checks, protected credential storage, and validated cache filenames.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (33)

Tainted flow: 'req' from os.environ.get (line 138, credential/environment) → urllib.request.urlopen (network output)

Critical
Category
Data Flow
Confidence
90% confidence
Finding

Credentials or environment variables flow to a network sink. This is a high-confidence indicator of credential exfiltration.

Content

Scanner excerpt · scripts/patentmax_client.py (reported line 142)May include surrounding context.

python
req.add_header("Content-Type", "application/json")
    req.add_header("Accept", "application/json")
    try:
        with urllib.request.urlopen(req, timeout=30) as response:
            payload = json.loads(response.read().decode("utf-8"))
    except urllib.error.HTTPError as exc:
        message = _server_message(exc.read().decode("utf-8", "replace"))

Tainted flow: 'req' from os.environ.get (line 333, credential/environment) → urllib.request.urlopen (network output)

Critical
Category
Data Flow
Confidence
94% confidence
Finding

The client attaches its API key to an unrestricted configurable base URL, including HTTP destinations, and uses urllib's default redirect handling, which can forward Authorization to another origin. An attacker-controlled configuration or a cross-origin redirect from the configured service can therefore expose a saved or environment-provided key; normal authentication to the default HTTPS API is otherwise legitimate.

Content

Scanner excerpt · scripts/patentmax_client.py (reported line 344)May include surrounding context.

python
req.add_header(key, value)

    try:
        with urllib.request.urlopen(req, timeout=TIMEOUT) as response:
            payload = response.read()
            # 服务端每个响应都回本次实扣和扣完后的余额。记下来,供 _spent/_balance 输出。
            # 拿服务端的真值,而不是在本地累加估价:跨进程累加迟早算岔,

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The supplied shell command exposes part of an API key rather than analyzing patent collaborations or generating a partner table. Although using an API credential could support the declared task, printing credential material is an undeclared behavior unrelated to that output.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The client can retrieve evidence useful for cooperation analysis, but the supplied code does not implement the central transformations or output promised by the description. Applicant statistics and candidate CSV export do not themselves produce a partner network with joint-patent counts, affiliate separation, topics, and activity assessments. API authentication, caching, and budget management are supporting implementation details and are not independently treated as mismatches. This assessment is limited to the supplied code chunk; an additional orchestration layer could potentially provide the missing workflow.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The supplied workflow explicitly starts from technical features of a proposal and expands results using the closest patent. It does not identify an organization, analyze joint applicants, separate affiliates, aggregate partnership metrics, or export a cooperation table. Patent retrieval could support the declared purpose, but this particular workflow is directed at finding technically related patents rather than analyzing collaboration networks.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

The supplied command searches for patents similar to a specific patent. It does not request co-filing partners or analyze collaboration relationships. Similarity search could support a broader patent workflow, but the supplied chunk establishes no connection to the declared partnership-analysis purpose.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
92% confidence
Finding

The supplied chunk requests patent claims for a hard-coded publication. It does not request co-applicant relationships, aggregate joint patents, distinguish affiliates, or produce a collaboration table. Although patent retrieval could support a broader analysis, this claims-specific request has no demonstrated connection to the declared collaboration workflow. The API authentication itself is not a mismatch.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
90% confidence
Finding

The supplied commands implement a single-patent research workflow, not the declared cooperation-network analysis. They do not visibly identify co-applicants, distinguish affiliates, aggregate joint filings, assess partnership activity, or export a relationship table. Patent retrieval could support the declared analysis, but this chunk does not establish that connection and instead focuses on citations and similarity. The client's internal behavior is not supplied, so the assessment is limited to the visible commands.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
90% confidence
Finding

CSV export is consistent with the declared output format, and candidate management could support a broader analysis workflow. However, the supplied chunk specifically implements candidate review and selection, including keep/exclude state changes based on technical relevance. It shows no partner identification, affiliate separation, joint-patent aggregation, or collaboration-activity analysis. The visible behavior therefore has a materially different immediate purpose. This assessment is limited to the supplied commands; the underlying client's implementation is not shown.

Content

No source excerpt is available for this finding.

Code capability 'env' detected in scripts/patentmax_client.py but not covered by declared permissions.

High
Category
MCP Least Privilege
Confidence
75% confidence
Finding

The skill uses 'env' capability that is not covered by the declared 'allowed-tools' preapproval. Review the host runtime's actual enforcement policy; runtime permissions are unknown unless separately verified, and this declaration alone does not establish a permission ceiling.

Content

No source excerpt is available for this finding.

Code capability 'network' detected in scripts/patentmax_client.py but not covered by declared permissions.

High
Category
MCP Least Privilege
Confidence
75% confidence
Finding

The skill uses 'network' capability that is not covered by the declared 'allowed-tools' preapproval. Review the host runtime's actual enforcement policy; runtime permissions are unknown unless separately verified, and this declaration alone does not establish a permission ceiling.

Content

No source excerpt is available for this finding.

The workflow sends user-supplied organization names and research queries to PatentMax without a privacy disclosure.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

The instructed statistics and search operations submit the user's target organization and partner-specific queries to a third-party patent service, potentially revealing research interests. The skill identifies PatentMax as its data source and discloses fees, but does not tell users that their search inputs will be transmitted to that service.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/api-reference.md (reported line 16)May include surrounding context.

bash
curl -s -H "Authorization: Bearer $PATENTMAX_API_KEY" \
  "https://api.ip930.com/api/patent?pn=CN109761224A"

服务端自己完成内部寻址,这些前置调用不进用户账单。

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/api-reference.md (reported line 48)May include surrounding context.

服务端直接返回上次的结果、不再扣费;同一个键换了参数会被拒(400);上一次还在处理中返回 409。

bash
curl -s -H "Authorization: Bearer $PATENTMAX_API_KEY" -H "Idempotency-Key: task42-search-1"   "https://api.ip930.com/api/search?q=%E6%B5%B8%E6%B2%A1%E5%BC%8F%E6%B6%B2%E5%86%B7&ds=cn&size=20"

网络超时、不确定上一次有没有成功时,用同一个键重发,不要换键。

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/api-reference.md (reported line 73)May include surrounding context.

curl -s -H "Authorization: Bearer $PATENTMAX_API_KEY"
--data-urlencode 'q=(固态电池 OR 全固态电池) AND 电解质'
--data-urlencode 'ds=all' --data-urlencode 'size=20'
-G "https://api.ip930.com/api/search"

text

> 检索式里有中文和括号,一定用 `--data-urlencode -G`,别手工拼 URL。

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/api-reference.md (reported line 89)May include surrounding context.

bash
curl -s -H "Authorization: Bearer $PATENTMAX_API_KEY" \
  "https://api.ip930.com/api/patent?pn=CN109761224A"

某一块上游取不到时,该字段为 null 并在 unavailable 里列出,其余照常返回;

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/api-reference.md (reported line 114)May include surrounding context.

bash
curl -s -H "Authorization: Bearer $PATENTMAX_API_KEY" \
  "https://api.ip930.com/api/claims?pn=CN109761224A"

只要同时需要两块以上,就用 /api/patent——它一口价,拆开单买反而贵。

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/faq.md (reported line 35)May include surrounding context.

md
- **密钥复制不全。** 前后有没有多余空格、有没有漏掉尾部字符
- **少了 `Bearer ` 前缀。** 用 curl 时 header 必须是 `Authorization: Bearer pm_live_xxx`,只写密钥不行
- **密钥被撤销了。** 到 [控制台](https://api.ip930.com/features/api-platform)看状态

> 还有个容易踩的:如果请求里带了 `Authorization` 头,URL 上的 `?apikey=` 就不会被读取。两种方式选一种,别混用——带了一个格式错误的 header,正确的 URL 参数也会被忽略。

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/faq.md (reported line 41)May include surrounding context.

md
## 402 余额不足

到[控制台](https://api.ip930.com/features/api-platform)充值。

新账户的体验额度分两笔:注册送 ¥10,第一次调用成功后再送 ¥40。如果第二笔一直没到,确认第一次调用确实返回了 200——失败的调用不算。

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/faq.md (reported line 119)May include surrounding context.

md
## 花了多少钱

脚本每次返回都带 `_spent`(本次实扣)和 `_balance`(余额),**这两个数来自服务端,是真值**。
准确的流水到[控制台](https://api.ip930.com/features/api-platform)查。

| 调用 | 单价 |
| --- | --- |

The workflow submits search terms derived from the user's technical invention to a paid patent-search service without warning about disclosure of potentially confidential information.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

The workflow directs the agent to construct external search queries from the user's technical subject, mechanism, and distinguishing features without first checking whether those details are confidential or authorized for disclosure. This could expose unpublished invention details or trade secrets to the search provider; routine searches using public applicant names present substantially less risk, and the provided file does not establish actual provider retention or misuse.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/tools.md (reported line 668)May include surrounding context.

试用用完、过期,或想用其他功能(查新、交底书等)时,命令会返回带 "free_trial": true 的错误,message 里写着下一步怎么办(注册 PatentMax 再送 ¥50 体验额度),原样转告用户,不要自己换写法重试。

用户有自己的密钥时用环境变量传,避免写进命令历史;设了就一律用它,不会领试用。密钥在 api.ip930.com 控制台创建:

bash
export PATENTMAX_API_KEY="pm_live_你的密钥"

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/api-reference.md (reported line 124)May include surrounding context.

健康检查

bash
curl -s "https://api.ip930.com/api/v1/health"

不需要密钥,用来确认网络通不通。

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · scripts/patentmax_client.py (reported line 102)May include surrounding context.

python
# 设了 PATENTMAX_API_KEY 就一律用它,不会去领试用。
TRIAL_SOURCE = "patent-search-cn"
TRIAL_GROUP = "patent-search"
TRIAL_REGISTER_URL = "https://api.ip930.com/features/api-platform"
KEY_FROM_TRIAL = False

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · scripts/patentmax_client.py (reported line 296)May include surrounding context.

python
if _billing["balance_cents"] < 500 and not KEY_FROM_TRIAL:  # 试用用户不提醒,用完时服务端会给注册指引
            out["_balance_warning"] = (
                "余额不足 ¥5,继续批量检索会很快耗尽。"
                "请先提醒用户到 https://api.ip930.com/features/api-platform 充值。"
            )
    if _billing["free_quota"]:
        left = max(0, _billing["free_quota"] - (_billing["free_used"] or 0))