T05 · Unauthorized Access and Privilege Escalation
- Location
SKILL.md:34- Finding
Unrestricted Search of Sensitive Historical Conversation Logs
- Content
View full analysis
/sessions/` (default: `~/.openclaw/agents//sessions/`). Use the `agent=` value from the system prompt Runtime line. - **`sessions.json`** - Index mapping session keys to session IDs - **`.jsonl`** - Full conversation transcript per session ## Structure Each `.jsonl` file contains messages with: ``` ```bash ### Extract user messages from a session jq -r 'select(.message.role == "user") | .message.content[]? | select(.type == "text") | .text' .jsonl ### Search for keyword in assistant responses jq -r 'select(.message.role == "assistant") | .message.content[]? | select(.type == "text") | .text' .jsonl | rg -i "keyword" ``` ```bash ### Search across ALL sessions for a phrase AGENT_ID="" SESSION_DIR="${OPENCLAW_STATE_DIR:-$HOME/.openclaw}/agents/$AGENT_ID/sessions" rg -l "phrase" "$SESSION_DIR"/*.jsonl ``` ### Technical Analysis The skill directs the agent to locate and inspect complete historical session transcripts. These files may contain user messages, assistant responses, tool results, personal information, credentials copied into prior conversations, and context unrelated to the current request. The documented commands allow both direct extraction of message content and unrestricted searches across every session belonging to the selected agent. The skill does not require explicit user confirmation, constrain access to a particular session or date range, redact sensitive values, or prevent unrelated historical content from being returned in the active conversation. This is a least-privilege concern rather than an operating-system privilege escalat ...[truncated 1867 chars]- Remediation
View remediation
