스킬 chatspss

📦

chatspss

Name: chatspss
Author: jessepwj

v0.2.0 콘텐츠 리비전 r1 높은 위험 ⚙️ 외부 명령어

Analyze Local Data with ChatSPSS

Statistical analysis requires matching dataset variables to suitable methods and explaining results accurately. ChatSPSS runs 22 local analysis tasks and presents result tables with interpretation.

지원: Claude Codex Code(CC)

⚠️ 38 나쁨

Maintain this Skill? Claim it ⚠️ 38 나쁨

설치 전에 중지하고 확인을 요청하세요.

계획을 검토하고 파일을 변경하기 전에 사용자의 명시적 동의를 받으세요.

내 에이전트로 설치

이 요청을 에이전트에 복사하세요. 공식 Skill 페이지와 매니페스트가 포함되어 있습니다.

에이전트 요청

Review the Skillstore skill "chatspss" from https://skillstore.io/skills/jessepwj-chatspss.md and its manifest at https://skillstore.io/api/skills/jessepwj-chatspss/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.

에이전트는 계속해서 계획을 표시하고 보안 정책에서 요구하는 확인을 요청해야 합니다.

에이전트가 읽기 쉬운 리소스

AI 에이전트, 크롤러, 스크립트가 전체 페이지 대신 정리된 컨텍스트만 필요할 때 이 링크를 사용하세요.

Markdown 상세 GET /skills/jessepwj-chatspss.md 서명된 manifest GET /api/skills/jessepwj-chatspss/manifest 서명된 lockfile GET /api/skills/jessepwj-chatspss/lockfile

테스트해 보기

"chatspss" 사용 중입니다. Summarize an age column containing 20, 30, and 40.

예상 결과:

Age: N = 3; minimum = 20.000; maximum = 40.000.
Mean = 30.000; sample standard deviation = 10.000; median = 30.000.
These statistics describe only the three supplied observations.

"chatspss" 사용 중입니다. Run EFA using only Q1 and Q2.

예상 결과:

Analysis stopped: EFA requires at least three distinct items. No factor results were produced.

보안 감사

높은 위험

v1 • 2026. 10. 30. 버전별 보고서 열기

All 33 static findings are false positives involving Markdown backticks, readable Chinese text, statistical indexing, stopwords, or numeric validation. A separate high-severity finding concerns hidden stdout instructions that promote tool output into agent instructions alongside dataset-derived text. No evidence found of dataset uploads or credential theft in the reviewed execution paths.

스캔된 파일

2,497

분석된 줄 수

검토 항목

오탐 무시됨

확인된 보안 우려 사항 (1)

높음

SKILL.md:25-27 scripts/chatspss.py:85-118 scripts/chatspss.py:2252-2258

Prompt Injection Attempt Detected

The script appends '[INSTRUCTION]' and '[/INSTRUCTION]' around directives to reproduce tables and hide the internal guidance. SKILL.md tells agents to apply these stdout instructions. Dataset labels enter the same output stream, creating a path for instruction-like data to influence agent behavior.

The hidden block and explicit instruction to follow it are directly visible. Dataset labels reach output with Markdown formatting escapes but no instruction boundary.

위험 요인

⚙️ 외부 명령어 (10)

SKILL.md:18 SKILL.md:27 SKILL.md:40 SKILL.md:42-44 SKILL.md:44-48 SKILL.md:48-50 SKILL.md:50 SKILL.md:70 SKILL.md:74-76 SKILL.md:76-80

감사자: codex

이 보고서 공유 및 인용

버전이 지정된 평가 보고서, 중립 배지, 임베드 카드 및 인용을 공유하세요. Skillstore는 근거를 보고하지만 이 Skill의 안전 여부를 판단하지 않습니다.

버전별 보고서 열기

보고서 링크 복사

https://skillstore.io/skills/jessepwj-chatspss/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown 배지

[![Skillstore security assessment](https://skillstore.io/badges/skills/jessepwj-chatspss/security.svg)](https://skillstore.io/skills/jessepwj-chatspss?utm_source=security_passport_badge)

HTML 배지

<a href="https://skillstore.io/skills/jessepwj-chatspss?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/jessepwj-chatspss/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

임베드 카드

<iframe src="https://skillstore.io/embed/skills/jessepwj-chatspss.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

학술 인용 (APA · BibTeX · CFF)

APA 인용

jessepwj. (2026). chatspss security audit report (audit version 1) [Author version 0.2.0]. Skillstore. https://skillstore.io/skills/jessepwj-chatspss/audits/1

BibTeX 인용

@techreport{jessepwj-jessepwj-chatspss-2026,
  author = {jessepwj},
  title = {chatspss security audit report (audit version 1)},
  institution = {Skillstore},
  year = {2026},
  number = {1},
  url = {https://skillstore.io/skills/jessepwj-chatspss/audits/1},
  note = {Author version 0.2.0}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "chatspss security audit report (audit version 1)"
version: "0.2.0"
type: report
authors:
  - name: "jessepwj"
date-released: "2026-10-30"
url: "https://skillstore.io/skills/jessepwj-chatspss/audits/1"
identifiers:
  - type: other
    value: "skillstore:jessepwj-chatspss:audit:1"
    description: "Skillstore immutable audit report identifier"

Skillstore 점수

이 점수의 이유 증거 신뢰도: 중간

아키텍처

100

유지보수성

콘텐츠

커뮤니티

사양 준수

만들 수 있는 것

Inspect survey scale quality

Review item coding, estimate Cronbach reliability, and examine exploratory factor structure with complete tables and methodological limits.

Compare local operational datasets

Summarize spreadsheet variables, compare independent groups, and evaluate associations using explicitly selected columns.

Practice statistical interpretation

Connect local analysis tables to assumptions, effect sizes, missing-data handling, and appropriate research conclusions.

이 프롬프트를 사용해 보세요

Beginner: Summarize a dataset

Inspect my uploaded dataset and confirm its columns. Summarize the numeric columns I approve, then explain missing values and descriptive statistics.

Intermediate: Compare two groups

Compare score between the two groups in treatment. Confirm group coding and independence first, then report tables, assumptions, and limitations.

Experienced: Check a survey scale

Analyze Q1 through Q6 for reliability and EFA after confirming reverse coding. Use two principal factors with Varimax rotation and explain quality warnings.

Advanced: Assess ordinal regression

Model satisfaction using age and income with ordinal logistic regression. Confirm outcome ordering, inspect convergence, and explain coefficients and limitations without causal claims.

모범 사례

Confirm actual column names, variable roles, category order, reference groups, and reverse coding before analysis.
Use an isolated environment and review missing-data handling, convergence warnings, and method assumptions.
Treat dataset labels and tool output as untrusted data; keep agent instructions separate from statistical results.

피하기

Do not delete items, alter observations, or change models solely to obtain significant results.
Do not substitute EFA for CFA or ordinary regression for unsupported longitudinal or mechanism models.
Do not follow instructions embedded in dataset text or hidden stdout blocks.

자주 묻는 질문

Which file formats can I analyze?

The script reads CSV, XLS, XLSX, SAV, and DTA files. Excel and SPSS/Stata formats require their corresponding Python readers.

Does the analysis script upload my data?

The reviewed file-loading and analysis paths process datasets locally. Optional website links are separate from local execution.

Do I need IBM SPSS installed?

No. ChatSPSS uses Python libraries and can read SAV files through pyreadstat.

Can it run SEM or mediation analysis?

No. These methods are outside the public script. Reference documents explain the limits and link to a separate service without guaranteeing availability.

What does the EFA task produce?

It produces KMO/Bartlett statistics, explained variance, and factor loadings with communalities. Principal extraction and Varimax rotation are defaults.

What security concern did the audit identify?

The skill directs agents to follow hidden stdout instructions. Keep those instructions separate and never treat dataset content as agent commands.