skill-researcher

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill instructs the agent to read and reproduce repository files (including YAML frontmatter, code blocks, and SKILL.md) and generate translated/copied outputs preserving original content, so any secrets present in those files would be included verbatim in the output.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill's required workflow (Step 2 "搜索热门项目" and Step 3 "读取每个项目的 SKILL.md") explicitly uses web search (WebSearch / mcp__web-search-prime__webSearchPrime) and repo-read tools (mcp__zread__get_repo_structure, mcp__zread__read_file) to fetch and analyze public GitHub repositories' SKILL.md and reference files, which are untrusted user-generated third-party content that can influence the agent's actions and recommendations.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill calls runtime repo-read tools (mcp__zread__read_file / mcp__zread__get_repo_structure) to fetch SKILL.md from external GitHub repositories (e.g., anthropics/skills/mcp-builder), and that fetched content is injected into the agent's prompts for analysis/translation and is required for the skill to operate.