Back to skill

Security audit

小说时间线约束核对

Security checks for vulnerabilities and agentic risk

Overview

The skill performs the promised novel timeline checks without hidden data access or system changes, though its optional installer is unpinned.

Use a reviewed local copy, or pin and review the installer and repository revision before installation. Supply only the events needed for the audit and manually review suggested chronology changes.

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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (7)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The supplied chunk is an installation command, not an implementation of the described audit. It can execute an external CLI and add a remotely sourced skill, while providing none of the declared analysis or outputs itself. The installed skill’s behavior cannot be verified from this snippet.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx skills'.

Medium
Category
MCP Rug Pull
Confidence
86% confidence
Finding

The installation command invokes the npm package skills through npx without pinning its version, potentially downloading and executing a changing package release with the user's privileges. A compromised release could therefore execute malicious code, although no malicious behavior is evident here. The finding's MCP label is inaccurate—this is an installation CLI—and the README requires manual execution rather than automatically running the command.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx skills'.

Medium
Category
MCP Rug Pull
Confidence
90% confidence
Finding

The installation command uses npx skills without pinning a version, allowing a changing npm package to execute with the invoking user's permissions. A compromised future release could execute malicious code during installation; the referenced skill repository is also not pinned to an immutable revision. This is an optional installation-time supply-chain risk, not an MCP-server reference or evidence of malicious intent, and the documented dependency-free local workflow reduces exposure.

Content

No source excerpt is available for this finding.

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

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 · scripts/self_test.py (reported line 9)May include surrounding context.

python
ROOT = Path(__file__).resolve().parents[1]
def run(name):
    return json.loads(subprocess.check_output([sys.executable, "-B", str(ROOT / "scripts/audit_timeline.py"), str(ROOT / "fixtures" / name)], text=True))

success = run("success-input.json")
failure = run("failure-input.json")

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 · scripts/self_test.py (reported line 104)May include surrounding context.

python
with tempfile.NamedTemporaryFile("w", encoding="utf-8", suffix=".json") as temp:
    json.dump({"events":[{"id":str(i),"scene":"S"+str(i),"time":"09:00"} for i in range(21)]}, temp)
    temp.flush()
    result = subprocess.run([sys.executable, "-B", str(ROOT / "scripts/audit_timeline.py"), temp.name], text=True, capture_output=True)
    assert result.returncode != 0 and "不得超过 20 个" in result.stderr
manifest = (ROOT / "manifest.yaml").read_text(encoding="utf-8")
skill = (ROOT / "SKILL.md").read_text(encoding="utf-8")

Scope Creep

Low
Category
Excessive Agency
Confidence
70% confidence
Finding

Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.

Content

Scanner excerpt · LICENSE (reported line 12)May include surrounding context.

text
permit persons to whom the Software is furnished to do so.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED,
INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A
PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE