semgrep-rule-creator

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly requires using WebFetch to read multiple public third‑party documentation links (e.g., https://semgrep.dev/docs/writing-rules/rule-syntax, https://appsec.guide/docs/static-analysis/semgrep/advanced/, and a public GitHub docs tree), so the agent will fetch and interpret untrusted external web content as part of its required workflow, which could materially influence subsequent actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's Documentation section explicitly requires using WebFetch at runtime to read external Semgrep documentation (e.g., https://semgrep.dev/docs/writing-rules/rule-syntax, https://appsec.guide/docs/static-analysis/semgrep/advanced/, https://semgrep.dev/docs/writing-rules/data-flow/constant-propagation, https://github.com/semgrep/semgrep-docs/tree/main/docs/writing-rules/), which the agent must fetch during execution and which directly guide its prompts/instructions, so these URLs are flagged.