> ## Documentation Index
> Fetch the complete documentation index at: https://smithers.sh/llms.txt
> Use this file to discover all available pages before exploring further.

# 0.25.6

> The output-task rollout reaches the rest of the init pack: every built-in workflow now ends with a deterministic output task, and bare smithers memory list lists facts across all namespaces instead of erroring.

Smithers 0.25.6 finishes what 0.25.5 started. In 0.25.5 the seeded workflows learned to end with an `output` task so a finished run prints a useful result. This release gives the remaining built-in (legacy-inline) workflows the same treatment, and fixes `bunx smithers-orchestrator memory list` so it works with no namespace.

***

## Workflows

Every built-in init-pack workflow now ends with a deterministic `output` task that aggregates its genuinely useful terminal result, so running one prints something meaningful instead of `output: null`:

* `plan`, `research`, `ticket-create`, and `tickets-create` surface the plan steps, key findings, ticket title, and ticket count they produced.
* `implement`, `debug`, `improve-test-coverage`, and `research-plan-implement` surface the files changed, whether validation passed, and whether a reviewer approved.
* `review` aggregates every reviewer's verdict into one outcome (reviewer count, approved, total and critical issue counts).
* `grill-me`, `feature-enum`, `audit`, and `workflow-skill` surface the resolution, feature counts, audited group count, and generated skill files.

The additions touch only the final output task; no existing task logic, ids, or agents changed, and all init-pack workflows still render. `vcs` and `mission` already end in deterministic aggregated result tasks (their action readers and `mission:final`), and `ralph` is an open-ended loop with no terminal node, so those three are unchanged by design.

## CLI

`bunx smithers-orchestrator memory list` now works with no namespace. It lists every fact across all namespaces, grouped by namespace, instead of failing with a raw validation error. Passing a namespace still lists just that namespace, exactly as before.

## Memory

The memory store gains a `listAllFacts()` method (and its `listAllFactsEffect` twin) that returns every fact across all namespaces, ordered by namespace then key. This backs the new no-namespace `memory list` and is available to any workflow that opens a memory store.

## Migration notes

No breaking changes are identified in 0.25.6. Workflows that already ended with their own `output` task are unaffected. If you copied a built-in workflow before this release and want the cleaner printed result, add a final `output` task that returns the aggregate you care about, the same pattern `hello` demonstrates.