coding-agent-sessions

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

• Third-party content exposure detected (low risk: 0.10). The runtime workflow scripts/find-agent-sessions.py search/find -> scripts/agent_sessions/cli.py (_search_payload/_get_payload) -> scanners* -> transcript/content_text ingests the user-input --query plus reads first/assistant user prompt text from local session transcript files (e.g., ~/.claude/.../*.jsonl, ~/.opencode/messages/ses_*/...) to match and summarize, so outsider-authored free text can enter only via compromised local stores but not via arbitrary external submission/queue.