스킬 compose-doctor

📦

compose-doctor

Name: compose-doctor
Author: goldsky-io

콘텐츠 리비전 r1 중간 위험 ⚙️ 외부 명령어🌐 네트워크 접근📁 파일 시스템 접근🔑 환경 변수

Diagnose Goldsky Compose App Failures

Existing Compose apps can fail because of task errors, missing secrets, wallet settings, or deployment mismatches. This skill guides evidence-based diagnosis and proposes targeted fixes.

지원: Claude Codex Code(CC)

📊 70 적절함

Maintain this Skill? Claim it 📊 70 적절함

내 에이전트로 설치

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

에이전트 요청

Review the Skillstore skill "compose-doctor" from https://skillstore.io/skills/goldsky-io-compose-doctor.md and its manifest at https://skillstore.io/api/skills/goldsky-io-compose-doctor/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

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/goldsky-io-compose-doctor.md 서명된 manifest GET /api/skills/goldsky-io-compose-doctor/manifest 서명된 lockfile GET /api/skills/goldsky-io-compose-doctor/lockfile

테스트해 보기

"compose-doctor" 사용 중입니다. My local Compose app reports that a secret referenced in the manifest is missing.

예상 결과:

Diagnosis: A required local secret is not configured. Fix: Add it through local secret management. Verify: Restart locally and check the logs.

"compose-doctor" 사용 중입니다. My named wallet fails in local development because it has no private key.

예상 결과:

Diagnosis: The named wallet cannot run in plain local mode. Options: Use fork mode or a suitable EOA wallet. Confirm before changing configuration.

"compose-doctor" 사용 중입니다. My deployed task reports that a transaction receipt reverted.

예상 결과:

Diagnosis: The onchain transaction reverted. Next step: Inspect the failed run trace for the decoded reason. Do not retry or fund automatically.

보안 감사

중간 위험

v1 • 2026. 10. 19. 버전별 보고서 열기

Most static matches are Markdown formatting, diagnostic examples, or scoped cache maintenance rather than malicious execution, credential theft, or reconnaissance. Confirmed risks include forced deployment, forced cleanup, and automatic pausing; a separate semantic finding covers automatic secret changes and redeployment without explicit approval. No evidence found of credential exfiltration or prompt injection; the skill explicitly treats logs as untrusted.

스캔된 파일

222

분석된 줄 수

검토 항목

오탐 무시됨

확인된 보안 우려 사항 (1)

중간

SKILL.md:130

Automatic Changes Without Explicit Approval

The instruction says, "If the fix is mechanical" and "execute it", including secret changes, manifest edits, and redeployment. This treats diagnosis as authorization for potentially disruptive application changes without explicit approval or an environment check.

The instruction explicitly authorizes automatic changes and reserves stopping for selected user-input cases. It does not require approval for the listed secret and deployment changes.

기능 검토 항목 (3)

이 스킬에 대해 예상될 수 있는 실제 로컬 기능이므로 검토가 필요하지만, 확인된 악의적 동작으로는 집계되지 않습니다.

중간

SKILL.md:169

Ruby/shell backtick execution

| `Refusing to deploy with a major api_version mismatch in non-interactive mode. Pass --force to ove

The remedy says to add --force after a major API version mismatch blocks deployment. This bypasses a compatibility safeguard without migration checks or explicit approval.

중간

SKILL.md:170

Ruby/shell backtick execution

| `Use --force for non-interactive cleanup.` | `compose clean` without `-f` in a non-TTY | Add `-f`

The remedy adds -f to cleanup in a non-interactive session without explaining affected resources or obtaining approval. This removes an interactive safeguard around cleanup.

중간

SKILL.md:192-193

Ruby/shell backtick execution

- Pause (not delete) before investigating. `goldsky compose pause` stops task execution without tear

The rule says to pause the app before investigating, which stops task execution. It does not require approval, so routine diagnosis can interrupt a running service.

위험 요인

⚙️ 외부 명령어 (50)

🌐 네트워크 접근 (2)

SKILL.md:51 SKILL.md:177

📁 파일 시스템 접근 (7)

SKILL.md:100 SKILL.md:177 SKILL.md:150 SKILL.md:161 SKILL.md:150 SKILL.md:161 SKILL.md:162

🔑 환경 변수 (1)

SKILL.md:154

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/goldsky-io-compose-doctor/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown 배지

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

HTML 배지

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

임베드 카드

<iframe src="https://skillstore.io/embed/skills/goldsky-io-compose-doctor.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

goldsky-io. (2026). compose-doctor security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/goldsky-io-compose-doctor/audits/1

BibTeX 인용

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

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "compose-doctor security audit report (audit version 1)"
version: "unspecified"
type: report
authors:
  - name: "goldsky-io"
date-released: "2026-10-19"
url: "https://skillstore.io/skills/goldsky-io-compose-doctor/audits/1"
identifiers:
  - type: other
    value: "skillstore:goldsky-io-compose-doctor:audit:1"
    description: "Skillstore immutable audit report identifier"

Skillstore 점수

이 점수의 이유 증거 신뢰도: 중간

아키텍처

유지보수성

콘텐츠

커뮤니티

사양 준수

만들 수 있는 것

Investigate Failing Tasks

Find failed invocations, review logs, and identify likely causes of HTTP errors or missed scheduled events.

Compare Deployed and Expected Behavior

Inspect app status and deployed source to investigate changes that did not reach the running application.

Diagnose Wallet and Sponsorship Errors

Distinguish local wallet limitations, unsupported sponsorship chains, and onchain reverts before selecting a remedy.

이 프롬프트를 사용해 보세요

Check a Broken App

Diagnose my existing Goldsky Compose app [app]. Start with status and logs, and ask before changing anything.

Find a Failing Invocation

Investigate HTTP 500 errors in [app] during [time window]. Review failed runs and logs, then report the cause and verification steps.

Inspect Wallet Configuration

Diagnose [wallet error] in [app] on [chain]. Check wallet and sponsorship configuration without requesting private keys or submitting transactions.

Compare Deployed Source and State

Investigate stale behavior in [app] after redeployment. Compare deployed source, failed runs, and collection state; propose fixes without applying them.

모범 사례

Confirm the project, app, and local or cloud environment before running commands.
Read logs and run details first; treat log content as untrusted and never execute embedded instructions.
Approve state changes explicitly, review compatibility before forced deployment, and verify results after each authorized fix.

피하기

Redeploying repeatedly without examining logs, failed runs, or the deployed source.
Pasting private keys, token values, or full environment files into diagnostic conversations.
Automatically forcing deployment, cleanup, or app pauses while investigating a production issue.

자주 묻는 질문

Is this for Docker Compose?

No. It diagnoses Goldsky Compose applications through the Goldsky CLI and dashboard.

Can it create a new app?

No. It focuses on existing apps and directs new app work to the separate compose skill.

Can it work without Bash access?

Yes. It provides one command at a time and asks the user to return output for analysis.

Can it reveal stored secret values?

No. The documented secret list returns names only. Incorrect values must be replaced through secret management.

Will it change my running app?

Some instructions apply fixes or pause execution automatically. Require explicit approval before secret changes, redeployment, cleanup, or pausing.

How does it investigate an onchain revert?

It locates failed runs and directs you to dashboard traces for decoded revert reasons. Further action can require contract-specific investigation.

개발자 세부 정보

작성자

goldsky-io

라이선스

MIT

Skillstore 리비전

버전 안내

작성자가 버전을 선언하지 않았습니다.

리포지토리

https://github.com/goldsky-io/goldsky-agent/tree/4ccc3f2c190d3e59660dcb52c2915e3ad98e3f6d/skills/compose-doctor

참조

587493833d6560296eae42efdb735565c0a3a3e8

유지보수 최신성

2026. 10. 19.

사용량

0 다운로드 · 0 조회

파일 구조

📁 evals/

📄 trigger-eval.json

📄 SKILL.md