Each model visit now gets <run>/stages/<seq>-<stage>/ with input.md,
memory.md, output.md, private/ and a shared run-level public/. The model
reads the two input files, follows the stage's instructions verbatim and
returns complete markdown; only the # outcome section is parsed.
- add internal/turn layout, render and output parsing
- write input.md/memory.md before a visit and output.md from the final
message; inventory all turn files as stage artifacts
- accumulate completed turns and system-written public file additions in
memory, including across retries and resume
- remove the JSON envelope, strict field validation, line format and the
repair ladder for model stages
- republish pipeline authoring prompts and docs in the new style
Checks: go build ./..., go vet ./..., go test -count=1 ./..., gofmt -l .
Drop the resource allocation feature end to end: store tables and migration, runner scheduling and cleanup, protocol types and commands, CLI commands, TUI screens, tests and docs. Existing databases open cleanly and definitions carrying a retired resources block still validate.
The operator's chat list showed ~99 stray "New chat" entries. The listing
was already fail-closed (registry-only); the fault was what got recorded: an
automated UI check drives the real page, so its throwaway chat was recorded as
an operator chat exactly like a manually created one and stayed in the list for
ever (each run added one).
Server: chat_create takes an optional ephemeral flag. The session is created,
bootstrapped, usable, closable and reopenable exactly like any other, but its
registry row is hidden, so chat_sessions never lists it. operator_sessions gains
an ephemeral column (idempotent migration, existing rows default to listed) and
the list query filters it; the single-row lookup does not, so close/reopen keep
working. RecordVerificationSession is the hidden recorder.
Web: the page marks a create ephemeral only when automation asks for it
(__pipelinesEphemeralChat, set by the UI smoke for its throwaway chat), so the
smoke's own runs stop polluting while a manual create is unchanged.
Tests: a hidden row is never listed yet is still looked up and closable; a
normal re-record of the same id lists it again; a pre-ephemeral database
migrates with every existing chat still listed. Live data pruned separately
(107 -> 8 registry rows, backed up).
The chat background-task bar now lists only tasks that are still running, so an
exited task no longer lingers and the bar disappears when the last one ends. The
one-line 'running now and recently observed exits' subtitle is replaced by a
compact 'Background tasks · N running' header, and the list is capped at three
touch-sized rows and scrolls beyond that instead of pushing the composer down the
screen. Unit-tested (running-only set and the row cap) and the stale note
updated.
The Android UI reviewer declares the android-devices pool instead of one
hard-coded Pixel, so a second UI run can use the idle foldable emulator instead
of waiting behind the single phone. Updates the Android map's device rules to
match the pool (replacing the old single-phone / no-emulator rule) and records
the pool, its members and the delivery-only personal Fold in a note. Safety
rules are unchanged; configuration and the pipeline version are separate steps.