zynkr

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

• Third-party content exposure detected (high risk: 0.75). /zynkr’s runtime workflow can ingest outsider-authored free text when it classifies an external-skill-url or pipeline-issue-ref and then fetches/reads the referenced GitHub issue/remote SKILL.md content (via the sub-skill calls like /skill-sourcer and /skill-qa), which is then placed into the agent’s LLM context for routing/QA/publishing.