Find Tasks and Read Sessions from the CLI
Published
Updated
Use wallfacer tasks list to find tasks across your active accounts, then wallfacer messages list to read a session. Sign in with wallfacer auth login. Use wallfacer auth status to check the configured account for account-specific commands.
Find a task across accounts
wallfacer tasks list returns tasks you own across your active accounts by default (--scope mine). Use --scope visible to include every task you can see in those accounts. Each result includes its account ID and name, so you can open the task in the right account.
wallfacer tasks list --q "login redirect"
wallfacer tasks list --scope visible --account-id <account-id> --q login
wallfacer tasks list --scope visible --account-ids <account-id>,<other-account-id> --q login
wallfacer tasks list --archived true--account-ids narrows the cross-account list to up to 100 comma-separated account IDs without changing the scope. IDs for accounts you cannot read are ignored; the result is empty if none are readable. Use --account-id for one account instead. It returns 403 or 404 when that account is inaccessible. Do not combine the two flags. --archived lists archived tasks instead of active tasks. Search matches titles, slugs, and prompts across the filtered set, including short or partial words. Results stay ordered by recent activity, even with --q.
Find tasks in the configured account
Use wallfacer tasks account-list for the account's task list and its status, owner, creator, and inbox filters:
wallfacer tasks account-list --q login --status done --owner-user-id <user-id> --per-page 20On this account list, --q searches titles, slugs, and prompts. Matches are ordered by relevance, then ID. Check the configured account with wallfacer auth status before using this command.
For either list, spell --q with two dashes. The short -q is the global --query JMESPath projection, which filters CLI output rather than searching tasks. Search accepts up to 255 characters.
Both lists return 25 tasks per page by default and accept --per-page up to 100. When meta.next_cursor is non-null, pass it as --cursor with the same filters. The cross-account list applies filters before paging and returns 422 if you reuse a cursor with changed filters or one issued before this behavior changed. Restart without a cursor in either case. You can change --per-page between pages:
wallfacer tasks list --scope visible --q login --per-page 20
wallfacer tasks list --scope visible --q login --per-page 20 --cursor <meta.next_cursor>A task with new activity while you page can move ahead of the cursor. Restart from the first page to see its new position.
Read a session as a conversation
List a task's sessions to find the session ID, then request text output:
wallfacer sessions list <task-id>
wallfacer messages list <task-id> <session-id> -o text --per-page 100The text view prints each turn with its speaker, role, time, and message ID. User turns also name their channel. Platform-authored rows are attributed to Platform. Turns appear in the order returned by the API.
By default, the text view omits tool calls, tool results, and transcript marker rows, and reports their counts in a Not shown: line. Add --tools to show tool calls and results, or --markers to show system, result, title, and session markers:
wallfacer messages list <task-id> <session-id> -o text --tools --markersThe text view omits thinking content and renders every image block as [image], including images returned in full by the API. Oversized tool inputs and results may appear as [elided: N bytes; ...] in the default trimmed view. Use JSON output with --view full, or fetch one message by ID with -o json, to inspect full image or thinking blocks:
wallfacer messages get <task-id> <session-id> <message-id> -o jsonWhen more messages exist, the text output includes a Next: line with the complete command for the next page. Run that command to continue. JSON remains the default for scripts; --tools and --markers require -o text.
Related
- Handbook, Team, and Playbooks from the CLI for starting a chat or playbook run.
- Tasks and Sessions for the task, session, and message API.