Back to skill

Security audit

收入确认·五步法逐条挂源

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marked this skill suspicious because scanners report concerns about external document parsing, undeclared network and local usage-tracking behavior, and instructions that allow calculations outside the stated scope.

Consider the reported scope mismatch, external document-parsing privacy concerns, and home-directory usage tracking before installing. The scan is partial because a referenced path was missing, so its evidence is incomplete; the verdict is suspicious, not malicious.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
Findings (28)

YARA rule 'agent_skill_destructive_autonomous_actions': Autonomous destructive filesystem, shell history, or repository actions in AI agent skills [agent_skills]

High
Category
YARA Match
Confidence
75% confidence
Finding

YARA rule matched a known malware signature (reverse shell, backdoor, ransomware, C2 framework, or info stealer).

Content

Scanner excerpt · README.en.md (reported line 148)May include surrounding context.

md
entry script drops an empty file at `~/.cue-usage/<slug>-<date>`.

**Why:** both platform download counters are unreliable — a new listing gets 50–60 seeded in its first 48 hours, and that is not users downloading. This local file is the only evidence that someone actually installed and ran this piece.

**It never gets in the way:** any failure (read-only home, full quota, no HOME) is skipped silently — no error, no exit-code change, no noise. Running it several times a day leaves one file.

To clear it: `rm -rf ~/.cue-usage`.

---

## 9. Files

| File | Purpose |
|---|---|
| `SKILL.md` / `SKILL.en.md` | Chinese and English skill description and check contract |
| `scripts/check_revenue.py` | five checks; `--selftest` runs the built-in fixtures, `--demo` prints a sample |
| `scripts/fixtures/` | one good sample + one bad sample + the three-tier runner |
| `tests/test_revenue.py` | package tests (entry contract, zero input, beacon) |

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
90% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · README.en.md (reported line 150)May include surrounding context.

md
**It never gets in the way:** any failure (read-only home, full quota, no HOME) is skipped silently — no error, no exit-code change, no noise. Running it several times a day leaves one file.

To clear it: `rm -rf ~/.cue-usage`.

---

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
90% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · README.md (reported line 150)May include surrounding context.

md
**它不会碍事:** 任何失败(家目录只读、配额满、无 HOME)都静默跳过,不报错、不改退码、不吐一句杂音。同一天跑多次只留一枚。

要清掉:`rm -rf ~/.cue-usage`。

---

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
90% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · README.en.md (reported line 150)May include surrounding context.

md
**It never gets in the way:** any failure (read-only home, full quota, no HOME) is skipped silently — no error, no exit-code change, no noise. Running it several times a day leaves one file.

To clear it: `rm -rf ~/.cue-usage`.

---

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
85% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · README.en.md (reported line 150)May include surrounding context.

md
**It never gets in the way:** any failure (read-only home, full quota, no HOME) is skipped silently — no error, no exit-code change, no noise. Running it several times a day leaves one file.

To clear it: `rm -rf ~/.cue-usage`.

---

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
85% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · README.md (reported line 150)May include surrounding context.

md
**它不会碍事:** 任何失败(家目录只读、配额满、无 HOME)都静默跳过,不报错、不改退码、不吐一句杂音。同一天跑多次只留一枚。

要清掉:`rm -rf ~/.cue-usage`。

---

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The reported home-directory usage markers are unrelated to producing a revenue-recognition register and are not explained in the user-facing workflow. If implemented as reported, they persist skill-use metadata without informed consent; the implementation is not included in this source, and there is no evidence of remote transmission.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

代码展示的是外部服务与工具发现,而非财报读取、收入确认分析或证据登记。声明未提及 MCP 服务集成,也未说明这些外部资源与收入确认分析存在关联,因此该网络访问与工具枚举不能据此视为已声明用途的支持性实现细节。代码中未见上传财报内容或执行所发现工具的行为。

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The supplied code performs remote service discovery and tool enumeration, not financial-report reading or evidence-backed revenue-recognition extraction. Although tool discovery can be a supporting implementation detail, this chunk establishes no connection between the queried services and the declared task. Its external resource access is therefore an undeclared capability in the supplied context. The code does not demonstrate financial-data exfiltration or execution of the discovered tools.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

This reports the same undisclosed persistent usage-marker behavior as another finding in this source. Such markers can reveal which skill was used and when, although the supplied skill document does not establish the implementation or any external collection.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

The description excludes calculating variable-consideration expected values, but the operational instructions permit self-calculation when an estimation method and inclusion limit are cited. This contradictory boundary can cause the agent to produce financial estimates that users reasonably expect the skill not to generate; source citations do not resolve that scope violation.

Content

No source excerpt is available for this finding.

Env Variable Harvesting

High
Category
Data Exfiltration
Confidence
60% confidence
Finding

Code enumerates, copies, or searches environment variables for secrets. Bulk environment access can collect credentials unrelated to the skill's stated purpose.

Content

Scanner excerpt · tests/test_revenue.py (reported line 31)May include surrounding context.

python
def run(*args, env=None, cwd=None):
    e = dict(os.environ, PYTHONDONTWRITEBYTECODE="1")
    if env:
        e.update(env)
    return subprocess.run([sys.executable, "-B", str(args[0]), *map(str, args[1:])],

Env Variable Harvesting

High
Category
Data Exfiltration
Confidence
60% confidence
Finding

Code enumerates, copies, or searches environment variables for secrets. Bulk environment access can collect credentials unrelated to the skill's stated purpose.

Content

Scanner excerpt · tests/test_revenue.py (reported line 134)May include surrounding context.

python
os.chmod(td, 0o700)

    def test_no_home_is_silent(self):
        e = dict(os.environ, PYTHONDONTWRITEBYTECODE="1")
        e.pop("HOME", None)
        r = subprocess.run([sys.executable, "-B", str(CHECK), "--demo"],
                           capture_output=True, text=True, env=e, cwd=str(PKG))

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 · README.en.md (reported line 53)May include surrounding context.

CATALOG="https://cuecue.cn/api/mcp-catalog" curl -s "$CATALOG" | head -c 2000

take this group's routing address from the catalog, then confirm the tools are present

curl -s -X POST "$ROUTING_URL"
-H 'Content-Type: application/json'
-d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

text

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 · README.md (reported line 53)May include surrounding context.

CATALOG="https://cuecue.cn/api/mcp-catalog" curl -s "$CATALOG" | head -c 2000

从目录里取本组的路由地址,确认工具在场

curl -s -X POST "$ROUTING_URL"
-H 'Content-Type: application/json'
-d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

text

The document-parsing workflow lacks a clear warning that user-supplied filing contents may be transmitted to an external service.

Medium
Category
Output Handling
Confidence
85% confidence
Finding

The workflow directs user-supplied documents into a parsing channel without clearly identifying whether their contents leave the workspace or requiring authorization for that transfer. Public filings reduce the usual sensitivity, but private drafts or annotated documents could be disclosed if externally parsed.

Content

No source excerpt is available for this finding.

The document-parsing workflow lacks a privacy warning about sending user-supplied files to an external service.

Medium
Category
Output Handling
Confidence
86% confidence
Finding

The instructions route user-provided PDFs and announcements through a parsing channel without clearly explaining external content transfer or obtaining authorization. Although annual reports are usually public, confidential drafts and annotations could cross the workspace boundary if that channel is external.

Content

No source excerpt is available for this finding.

The workflow permits computing variable consideration using the expected-value method, despite the manifest explicitly excluding that service.

Medium
Category
Excessive Agency
Confidence
97% confidence
Finding

The workflow explicitly permits computed variable consideration using an expected-value method, despite advertising that service as excluded. This contradictory boundary can lead an agent to produce accounting estimates outside the user's selected scope; labeling them as computed and requiring sources reduces, but does not eliminate, the integrity risk.

Content

No source excerpt is available for this finding.

Attachment parsing lacks a warning that document content will be sent to an external service.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

Attachment parsing is triggered by where a quotation resides rather than by authorization to disclose the attachment to an external parser. Service and billing references do not clearly communicate the content-transfer boundary, creating a confidentiality risk for private or annotated financial documents.

Content

No source excerpt is available for this finding.

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: env, file_read, file_write, shell.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

The workflow permits expected-value calculations of variable consideration that the manifest explicitly excludes.

Medium
Category
Excessive Agency
Confidence
96% confidence
Finding

The workflow allows self-calculated variable consideration using an expected-value estimation method despite the manifest's explicit exclusion. This creates an output-integrity and scope-control failure in a financial workflow, even when calculations are labeled and sourced.

Content

No source excerpt is available for this finding.

The workflow sends attachments to an external parsing service without warning that user-provided document content will leave the local environment.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The workflow directs attachments and scans to an external parsing service without requiring confirmation that user-provided content may leave the local environment. Public filings reduce the risk, but private or annotated attachments could disclose confidential information without informed authorization.

Content

No source excerpt is available for this finding.

L129-L143 implement persistent usage tracking unrelated to revenue-recognition analysis.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The checker unconditionally creates a dated usage marker under ~/.cue-usage before performing its requested work, including demo and self-test operations. This is unnecessary persistent state outside the workspace and can expose local execution history; the code shows no network reporting or document-content collection.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · tests/test_revenue.py (reported line 34)May include surrounding context.

python
e = dict(os.environ, PYTHONDONTWRITEBYTECODE="1")
    if env:
        e.update(env)
    return subprocess.run([sys.executable, "-B", str(args[0]), *map(str, args[1:])],
                          capture_output=True, text=True, env=e, cwd=str(cwd or PKG))

The tests require persistent usage tracking unrelated to revenue-recognition analysis.

Medium
Category
Excessive Agency
Confidence
95% confidence
Finding

These tests explicitly require a persistent, skill-specific daily usage marker, making unrelated usage recording a required product behavior. The shown assertions use temporary HOME directories and empty files, so they demonstrate a privacy-relevant design requirement rather than remote telemetry or sensitive-content extraction.

Content

No source excerpt is available for this finding.