infer-upgrade

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly runs and echoes the contents of ~/.infer/config.json (MCP_CONFIG=$(cat ~/.infer/config.json ...)) which can contain API keys or tokens and would cause the agent to read and output secret values verbatim.

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly queries public package metadata (e.g., "npm view @inferevents/sdk") and re-installs/fetches code via npx and "npx skills add infer-events/skills" which pulls latest code from GitHub, so untrusted third-party package content and repository code are fetched and can influence update decisions and runtime behavior.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill runs npx to fetch and execute remote packages at runtime (e.g., npx --yes @inferevents/mcp@latest and npx skills add infer-events/skills which pulls from GitHub), so external code is fetched and executed that the skill depends on.