Automations Every run is inspectable.

/automations/runs Available now

No “it ran” mystery. Every execution leaves a persistent record, and you control who is told when a run fails.

[ 02 // run history & observability ]
[ 02 ]

Every run is inspectable.

No “it ran” mystery. Every execution gets a persistent record with status, timing, tokens, tool calls, credits, summary, and a transcript you can pull on demand.

One automation run: its duration, tool calls, cost and tokens, the tool calls in order marked read or write, and a summary of what it did.
One run: what it cost, each tool it called in order, read or write, and what it did. Ticket details replaced.
morning-ticket-review
sched · 0 7 * * 1-5 · America/Chicago last 5 runs
StartedStatusDurationToolsTokensCreditsSummary
Nov 13 07:00✓ Completed47s916,384712 priority, 3 SLA risk
Nov 12 07:00✓ Completed52s1118,120815 priority, 1 client impact
Nov 11 07:00! Timed out120s1422,00410Hit max runtime, partial summary
Nov 10 07:00✓ Completed41s814,29069 priority, clean backlog
Nov 9 07:00✓ Completed39s713,880611 priority, 2 new escalations
Completed Running Pending Timed out Out of credits Failed Interrupted

Failure emails

Switch on Email me when a run fails, per automation, and pick the address.

Versions and rollback

Every save is recorded as a version. Restore any earlier version as the live setup.

Memory

Switch on Remember across runs and the automation keeps notes between runs. It is off by default.

Run as

Choose whose connector access an automation uses: a service account bound to that automation, or a named member. If that credential is missing, the run is refused.

Owners and administrators can delete a run’s transcript, or set it to delete itself when the run ends.

[ 03 // a sample run ]
[ 03 ]

One run, as it is logged.

An illustration of a run log: what was checked, what was called, what it cost and how it ended.

run log · morning-ticket-review · run 01KX…
[pre-flight]  tenant ok · consent ok · sync=Synced · capacity ok
[credits]     reserved 25 cr · cap 50 cr · balance 325 cr
[mcp]         scope = [halo_list_tickets, halo_get_ticket, halo_list_assets, …]  (7 tools)
[run]         started 2026-11-13T07:00:02-06:00 · model claude-sonnet-5-5
→ halo_list_tickets(status=open, priority=≤2, limit=50)
→ halo_get_ticket(id=48231) · halo_get_ticket(id=48244) · …
→ halo_list_assets(client_id=812)
   [summary] 12 priority tickets · 3 SLA risk · 1 client impact …
[tokens]      in 14,204 · out 2,180 · cache-r 8,992
[tools]       9 calls · runtime 47s
[credits]     finalized 7 cr · refunded 18 cr · balance 318 cr
✓ Completed · run id 01KX… · view transcript_

1 credit is $0.10 of usage today.

[ 04 // faq ]
[ 04 ]

Questions.

Can I see what an agent did during a run?

Yes. Every run has a record with status, timing, token counts, tool-call count, credits consumed, a summary, trigger payload, and a full transcript you can fetch on demand.

Does StackJack see or store my data?

StackJack does not keep a general copy of your business data. During a run, your connectors’ answers pass through StackJack to the AI model. Each run keeps a record: status, timing, token counts, cost, a summary, the trigger payload and the inputs sent to each tool. The conversation, which can include tool results, is kept so you can review the run; for a run on credits or on your own Anthropic key, Anthropic holds the session too, and a run on your Azure Foundry has no session at Anthropic. Owners and administrators can delete a run’s transcript at any time, or set transcripts to delete themselves when a run ends. Deleting removes the conversation from StackJack and from our AI provider. See the full security brief.

[ 06 // start ]

Recurring work, managed.

Automations are available today for every StackJack user. Run them on your Azure Foundry or your own Anthropic key for a flat monthly rate, or buy a credit pack to try one. Packs never expire, and you see real-time usage against your balance as your agents run.