Reference
Command reference
Everything the CLI accepts. Flags compose unless noted.
Filters
| Flag | Effect |
|---|---|
--days N | Only the last N days. The fastest way to make a slow run fast. |
--project SUBSTR | Only projects whose name contains SUBSTR. |
--agent claude|codex|all | Which agents to include. Default all. |
--top N | Projects to list. Default 12. |
--root DIR | Read transcripts from DIR instead of the default locations. |
Views
| Flag | Effect |
|---|---|
--bash | Shell audit only. Skips cost entirely. |
--coach | Coaching findings only. |
--agents | Which agent platforms are installed, and whether their binaries are validly signed by the expected publisher. |
Explaining a number
| Flag | Effect |
|---|---|
--explain | List the topics you can ask about. |
--explain TOPIC | How a number is computed, what it assumes, and how to check it. Topics: sources, cost, cache, tickets, secrets, subagents, shell, coach, agents. |
--why AFxxx | Why one coach finding fired, against your actual numbers. |
Output
| Flag | Effect |
|---|---|
--json | Machine-readable output. Redacted by default. See integrations. |
--share | A postable summary containing nothing identifying — no project names, branches, paths, commands or identifiers. |
--no-redact | Do not redact credentials from output. Unsafe to share or pipe to a file you forget about. |
Live monitoring
| Flag | Effect |
|---|---|
--watch | Poll for new secrets and risky commands, notifying as they appear. |
--interval SEC | Poll interval. Default 4. |
--quiet | With --watch: notify on secrets only, not every flagged command. |
Other
| Flag | Effect |
|---|---|
--mcp | Run as an MCP server over stdio so an agent can query itself. |
--version | Print the version. |