SkillForge API 服务发现

AdvisoryAudited by Static analysis on Apr 30, 2026.

Overview

No suspicious patterns detected.

Findings (0)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

NoteMedium Confidence
ASI01: Agent Goal Hijack
What this means

A malicious or compromised marketplace listing could include persuasive text that tries to influence the agent or user.

Why it was flagged

Service names/descriptions/developer text returned by the external marketplace are inserted into the agent-facing output. This is purpose-aligned, but that retrieved text should be treated as untrusted data rather than instructions.

Skill content
lines.push(`   ${service.description}`); lines.push(`   开发者: ${service.developer || '匿名'}\n`);
Recommendation

Treat service listings as informational only; confirm the selected provider and ignore any instructions embedded in service descriptions.

What this means

If invoked unintentionally, the skill could cause paid API calls or send user-provided input to an external service.

Why it was flagged

The skill explicitly supports invoking paid external services. This is its stated purpose and is disclosed, but it is a high-impact action because it can spend account balance.

Skill content
3. **调用服务** - 用户确认后调用服务并返回结果 ... 所有服务调用都会从账户余额扣除费用
Recommendation

Require explicit confirmation before any paid invocation, keep a low max-cost setting, and review the selected service and price before proceeding.

What this means

The API key may allow billed calls against the user's SkillForge account if mishandled or over-permitted.

Why it was flagged

The skill authenticates to SkillForge using a bearer API key, which is expected for service access and billing.

Skill content
'Authorization': `Bearer ${config.api_key}`
Recommendation

Use a dedicated/revocable API key with spending limits if available, store it securely, and prefer HTTPS SkillForge endpoints.

What this means

Prompts, files, or other inputs supplied for service invocation may be processed by external services rather than only locally.

Why it was flagged

The skill is designed to communicate with SkillForge and developer-provided services. This is disclosed and purpose-aligned, but it means user inputs may cross a third-party provider boundary.

Skill content
permissions:\n  - name: network\n    reason: 需要访问 SkillForge API 和开发者服务
Recommendation

Do not send sensitive or regulated data unless you trust the selected service and understand its retention/privacy terms.