mxr send
Generated from
mxr send --help. After editingcrates/daemon/src/cli/, update and review the snapshots withscripts/cargo-test -p mxr --test integration cli_help::, then rebuildsite/.
Send a draft by ID
mxr send [OPTIONS] <DRAFT_ID>Arguments
Section titled “Arguments”| Flag | Description |
|---|---|
<DRAFT_ID> |
Draft ID to send |
Options
Section titled “Options”| Flag | Description |
|---|---|
--expected-workspace <EXPECTED_WORKSPACE> |
Workspace JSON returned by the reviewed preview; required for sending |
--expected-revision <EXPECTED_REVISION> |
Revision returned by the reviewed preview. Required for sending now |
--operation-id <OPERATION_ID> |
Stable identity retained before submission; defaults to draft ID/revision |
--account <ACCOUNT> |
|
--dry-run |
Show what would be sent (sender, recipients, subject, byte count) without sending |
--format <FORMAT> |
[possible values: table, json, jsonl, csv, ids] |
--at <TIME> |
Schedule the draft to be sent later instead of sending now. Same forms as mxr snooze --until: in 2h, tomorrow 9am, monday 17:00, RFC3339. Use mxr unsend <draft-id> to cancel |
--remind-after <TIME> |
After sending now, remind if no reply has arrived by this time. Same forms as mxr remind --when: in 2h, tomorrow 9am, monday 17:00, RFC3339 |
--check |
Run the safety pipeline against the draft and exit without sending. Exit non-zero only if a Blocker issue is present |
--override-safety <TOKEN> |
Single-use override token (issued by a previous failed --check) to bypass a Blocker |
--no-llm |
Skip LLM-backed safety checks (answer-coverage). Useful when the daemon’s LLM is a rate-limited cloud model. Only honored with --check |
-h, --help |
Print help |
Use When
Section titled “Use When”Preview a saved draft, then send or schedule the reviewed revision. Keep the expected_workspace JSON from that original preview and pass it unchanged, shell-quoted, with --expected-workspace on send, schedule and outcome queries. For an immediate send, keep the preview operation ID and query it with mxr send-outcome after a lost response; unknown is not permission to retry. The scheduler assigns a separate identity: query its latest outcome with mxr send-outcome "draft:DRAFT_ID" --account ACCOUNT_ID --expected-workspace '<OWNER_JSON_FROM_SCHEDULE_PREVIEW>'. Add --check to run the pre-send safety pipeline without sending; exit 2 on Blocker. Use --override-safety <TOKEN> with the reviewed revision, original workspace and operation ID to bypass a Blocker the previous --check minted.
Everyday Examples
Section titled “Everyday Examples”mxr send DRAFT_ID --dry-run --format json # retain the revision, operation ID and expected_workspace JSONmxr send DRAFT_ID --expected-revision REVISION_FROM_PREVIEW --expected-workspace '<OWNER_JSON_FROM_PREVIEW>' --operation-id OPERATION_ID_FROM_PREVIEWmxr send DRAFT_ID --at 'tomorrow 9am' --expected-revision REVISION_FROM_PREVIEW --expected-workspace '<OWNER_JSON_FROM_PREVIEW>'mxr send DRAFT_ID --check --format json # safety report, no sendmxr send DRAFT_ID --expected-revision REVISION_FROM_PREVIEW --expected-workspace '<OWNER_JSON_FROM_PREVIEW>' --operation-id OPERATION_ID_FROM_PREVIEW --override-safety OVERRIDE_TOKEN_FROM_CHECK # after preview and --checkSee also
Section titled “See also”- CLI overview: full command index
- Concepts: query operators, search modes, JSON shapes
- Automation contract: which commands support
--format json,--dry-run, stdin