investigating-logs

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill instructs the agent to fetch and include raw log rows (1–2 sample raw lines) via posthog:query-logs, which can contain API keys, tokens, or passwords and provides no guidance to redact them, so the LLM may be required to output secrets verbatim.

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

• Third-party content exposure detected (medium risk: 0.30). SKILL.md describes a runtime workflow that calls PostHog log-analysis tools like posthog:logs-patterns(-diff), posthog:logs-count-ranges, and posthog:query-logs, which ingest free-form log message text that can be outsider-authored (e.g., user input written to logs) into the LLM.