diff --git a/README.md b/README.md
index 285f7a5..a32493b 100644
--- a/README.md
+++ b/README.md
@@ -80,6 +80,10 @@ jm read <email-id> --thread
 # Search for emails
 jm search "meeting agenda"
 jm search --from alice@example.com --after 2026-01-01T00:00:00Z
+jm search --from alice@example.com --after 2026-01-01
+
+# Mark emails as read
+jm mark-read <email-id>
 
 # Archive emails
 jm archive <email-id>
@@ -106,6 +110,7 @@ jm move <email-id-1> <email-id-2> --to Receipts
 | ------------------------------------- | ------------------------------------ |
 | `jm archive <id> [id...]`             | Move emails to the Archive mailbox   |
 | `jm spam <id> [id...]`                | Move emails to the Junk/Spam mailbox |
+| `jm mark-read <id> [id...]`           | Mark emails as read                  |
 | `jm move <id> [id...] --to <mailbox>` | Move emails to a specified mailbox   |
 
 See [docs/CLI-REFERENCE.md](docs/CLI-REFERENCE.md) for full details on all flags, output schemas, and examples.
diff --git a/docs/CLAUDE-CODE-GUIDE.md b/docs/CLAUDE-CODE-GUIDE.md
index 5e9d2d5..2d8546e 100644
--- a/docs/CLAUDE-CODE-GUIDE.md
+++ b/docs/CLAUDE-CODE-GUIDE.md
@@ -66,15 +66,16 @@ Add this to your project's `CLAUDE.md` to give Claude Code context about `jm`:
 
 - `jm archive <id> [id...]` -- move to Archive
 - `jm spam <id> [id...]` -- move to Junk
+- `jm mark-read <id> [id...]` -- mark as read
 - `jm move <id> [id...] --to <mailbox>` -- move to a named mailbox
 
 ### Notes
 
 - Output is JSON by default; errors are JSON on stderr with exit code 1
-- Email IDs from `list` and `search` chain directly into `read`, `archive`, `spam`, and `move`
+- Email IDs from `list` and `search` chain directly into `read`, `archive`, `spam`, `mark-read`, and `move`
 - Sending and deleting email are structurally disallowed
-- Date filters use RFC 3339 format (e.g., `2026-01-15T00:00:00Z`)
-- Batch operations (`archive`, `spam`, `move`) accept multiple IDs
+- Date filters accept RFC 3339 (e.g., `2026-01-15T00:00:00Z`) or bare dates (e.g., `2026-01-15`)
+- Batch operations (`archive`, `spam`, `mark-read`, `move`) accept multiple IDs
 ```
 
 ## Workflows
@@ -141,10 +142,10 @@ jm search --has-attachment --after 2026-01-01T00:00:00Z --before 2026-02-01T00:0
 
 ## Tips
 
-- **Chaining IDs:** Email IDs from `list` and `search` results chain directly into `read`, `archive`, `spam`, and `move`.
-- **Batch operations:** `archive`, `spam`, and `move` accept multiple email IDs in a single call.
+- **Chaining IDs:** Email IDs from `list` and `search` results chain directly into `read`, `archive`, `spam`, `mark-read`, and `move`.
+- **Batch operations:** `archive`, `spam`, `mark-read`, and `move` accept multiple email IDs in a single call.
 - **Filter-only search:** Omit the query argument and use only flags to search by sender, date range, attachments, etc.
-- **Date format:** All date flags (`--before`, `--after`) use RFC 3339 format, e.g. `2026-01-15T00:00:00Z`.
+- **Date format:** All date flags (`--before`, `--after`) accept RFC 3339 format (e.g. `2026-01-15T00:00:00Z`) or bare dates (e.g. `2026-01-15`).
 - **Thread view:** Use `jm read <id> --thread` to see the full conversation context for a single email.
 - **Mailbox names:** Both `--mailbox` and `--to` accept mailbox names (e.g., "Inbox", "Receipts") or mailbox IDs.
 - **Sort order:** `jm list --sort "subject asc"` sorts by subject ascending. Fields: `receivedAt`, `sentAt`, `from`, `subject`.