# Log > Show the SafeRun Guard audit log — what actions the agent executed in recent sessions Use when this capability is needed. - Skill: `tomevault-io/log` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/log` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/log/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: AI & ML - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/log --- Read the audit log file at `~/.saferun/audit.jsonl`. Each line is a JSON record of a tool action that was allowed and executed. Fields: `ts` (timestamp), `session` (session ID), `tool` (Bash/Write/Edit/Read), `input` (command or file path), `cwd` (working directory). Show the most recent 20 entries by default. If the user asks for more, show more. Format as a readable table or summary. Group by session if there are multiple sessions. If the file doesn't exist, tell the user no actions have been logged yet. --- > Converted and distributed by [TomeVault](https://tomevault.io/claim/cocabadger) — claim your Tome and manage your conversions.