The Agent Skills Directory

[SAFE]: The skill facilitates the analysis of local session logs located in the user's home directory (~/.openclaw/agents/). This access is consistent with its stated purpose of providing conversation history to the agent. No network operations or unauthorized file system modifications were detected.
[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by ingesting and processing historical conversation transcripts that may contain untrusted content from past interactions.
Ingestion points: Accesses and parses .jsonl files within the ~/.openclaw/agents/<agentId>/sessions/ directory.
Boundary markers: Absent; the provided shell commands extract raw message content using jq without wrapping the data in protective delimiters or applying "ignore" instructions.
Capability inventory: Uses shell scripting for file iteration and utilizes jq, rg, ls, awk, head, and tail for data processing.
Sanitization: No sanitization, filtering, or validation is applied to the historical text before it is presented to the agent for analysis.

session-logs