project-memory

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

• Third-party content exposure detected (medium risk: 0.65). This skill’s required load command reads the project’s persisted ~/.mystack/projects/<slug>/learnings.jsonl (a plain JSONL file whose lines can be written by prior/outside agents or processes) and prints its insight/files text into the LLM context via command output.