Skill Trust Decision
dlazy-webtoon-adapter
Skill is a legitimate webtoon-adaptation tool but declares no capabilities while the skill card mentions shell command output and the skill instructs agents to execute external CLI commands — capability declaration gap warrants caution.
Most direct threat evidence
Why this conclusion was reached
1/4 dimensions flagged Block
Declared vs actual capability
4 undeclared or violating capabilities were inferred.
Review
Hidden execution and egress
5 lower-risk artifacts were extracted and still need context.
Pass
Attack chain and severe findings
There is no explicit malicious chain in the report.
Review
Dependencies and supply chain hygiene
1 dependency or supply-chain issues need attention.
What drove the risk score up
Undeclared shell execution +20
SKILL.md instructs the agent to execute terminal commands (dlazy seedream-4.5) but declares no shell:WRITE capability
Undeclared network access +15
Skill communicates with api.dlazy.com and files.dlazy.com but does not declare network:READ in SKILL.md
Undeclared credential storage +10
Saves dLazy API key to ~/.dlazy/config.json without declaring filesystem:WRITE
Version inconsistency +5
SKILL.md frontmatter says 1.3.5, _meta.json says 1.3.25, install command says @1.2.3 and @1.0.5 — no consistent pinning
Most important evidence
Medium Doc Mismatch
Shell execution capability not declared in SKILL.md
The skill instructs agents to execute terminal commands (dlazy seedream-4.5) and skill-card.md lists 'Shell commands' as an output type, but SKILL.md declares no shell:WRITE capability. This is a doc-to-code mismatch.
SKILL.md:199 Add 'shell:WRITE' to the declared allowed-tools in SKILL.md frontmatter or metadata
Medium Doc Mismatch
Network access not declared in SKILL.md
The skill metadata lists api.dlazy.com and files.dlazy.com as API endpoints and the skill uploads media files to cloud storage, but no network:READ or network:WRITE is declared in SKILL.md.
SKILL.md:17 Add 'network:READ' (and network:WRITE for uploads) to declared capabilities in SKILL.md
Medium Priv Escalation
Credential persistence to local filesystem not declared
The skill instructs users to run 'dlazy login' or 'dlazy auth set' which writes API credentials to ~/.dlazy/config.json. This filesystem WRITE is not declared in SKILL.md.
SKILL.md:36 Declare filesystem:WRITE in allowed-tools for credential storage path
Low Supply Chain
Inconsistent and unpinned @dlazy/cli version across metadata
SKILL.md frontmatter says @1.2.3, webtoon-adapter metadata says @1.0.5, SKILL-cn.md body text says @1.2.3, skill-card.md says install field. No consistent version pinning across all fields increases supply-chain risk.
SKILL.md:6 Pin to a single consistent version across all metadata fields
Declared capability vs actual capability
Network Block
Declared NONE
→ Inferred READ
SKILL.md metadata: apiEndpoints: ["api.dlazy.com","files.dlazy.com"] Shell Block
Declared NONE
→ Inferred WRITE
SKILL.md execution section: 'dlazy seedream-4.5 --prompt "..."', skill-card.md: Output Type(s): Shell commands Filesystem Block
Declared NONE
→ Inferred WRITE
SKILL.md auth section: 'CLI 会把 key 保存在 ~/.dlazy/config.json' Environment Block
Declared NONE
→ Inferred READ
SKILL.md: 'DLAZY_API_KEY environment variable' — implies reading env vars Suspicious artifacts and egress
Medium External URL
https://www.npmjs.com/package/@dlazy/cli SKILL-cn.md:7
Medium External URL
https://dlazy.com SKILL-cn.md:32
Medium External URL
https://dlazy.com/dashboard/organization/api-key SKILL-cn.md:33
Medium External URL
https://clawhub.ai/user/dlazyai skill-card.md:9
Medium External URL
https://clawhub.ai/dlazyai/skills/dlazy-webtoon-adapter skill-card.md:39
Dependencies and supply chain
| Package | Version | Source | Known vuln | Notes |
|---|---|---|---|---|
| @dlazy/cli | * | npm | No | Versions vary across metadata fields (1.0.5, 1.2.3) — not consistently pinned |
File composition
4 files · 484 lines
Markdown 3 files · 479 linesJSON 1 files · 5 lines
Files of concern · 3
SKILL-cn.md https://www.npmjs.com/package/@dlazy/cli · https://dlazy.com · https://dlazy.com/dashboard/organization/api-key
SKILL.md Shell execution capability not declared in SKILL.md · Network access not declared in SKILL.md · Credential persistence to local filesystem not declared · Inconsistent and unpinned @dlazy/cli version across metadata
skill-card.md https://clawhub.ai/user/dlazyai · https://clawhub.ai/dlazyai/skills/dlazy-webtoon-adapter
Other files · _meta.json
Security positives
No malicious code (base64, eval, reverse shell) found — this is a documentation-only skill with no executable scripts
External CLI source code is publicly referenced (github.com/dlazy-ai/cli)
API key can be passed via DLAZY_API_KEY env var instead of being persisted locally
Skill-card.md correctly lists known risks and mitigations for external CLI and credential storage
skill-card.md explicitly recommends reviewing the npm package before installation