Add CapabilityAwareRoutingStrategy — it hard-filters to providers that declare every
required capability (like FirstAvailable) but ranks the matches by summed required-capability
score; ties keep list order, so it is a strict superset of first-available. The server now
wires it as the routing policy.
The orchestrator augments a stage's required capabilities with ModelCapability.ToolCalling
when the stage grants tools, so tool-heavy stages route to the best tool-calling model rather
than whichever healthy provider comes first.
Scores come from each provider's declared capabilities(); observed per-model reliability
(GET /metrics/tool-reliability) can feed in later as an additional weight.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Enable autonomous QA through a remote OpenAI-compatible provider (NVIDIA NIM)
and harden the tool/approval path so unattended multi-stage runs complete.
- inference: add openai_compat provider (Bearer chat-completions for NIM/OpenAI),
dispatched by provider type "nim"/"openai"; key via api_key/api_key_env.
- server: bind configured [server] host/port instead of a hardcoded 8080;
POST /sessions accepts an optional `intent` (WS parity) for intent-driven workflows.
- kernel: thread the bound operator profile's approval_mode into per-tool gating so
auto/yolo enable unattended approval (engine still consulted; policy/plane-2 BLOCK
stays terminal); on a recoverable tool failure feed the tool's arg-schema back into
context so the model self-corrects instead of repeating a malformed call.
- tools: split deletion out of file_write into a separate, explicitly-named file_delete
tool — a model can no longer delete a file by getting a write-mode parameter wrong.
- server: add GET /metrics/tool-reliability — per-model tool-call validity from the
event log (measurement groundwork for capability-aware routing).
- docs: update AGENTS.md across kernel, tools, server, inference.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
When a write is blocked for being outside the claimed task's
affected_paths, the implementer can propose widening scope via a new
T2 propose_scope tool. The operator decides (invariant #4 — the agent
can't self-grant). Approve unions the proposed paths into the task's
affected_paths (existing TaskAffectedPathsSetEvent); activeScope
re-derives the wider manifest from events so the same write then passes.
Reject blocks the task (via TaskClaimCoordinator.blockActiveTask, hooked
on both approval-denial branches, scoped to propose_scope) so the loop
advances. No new event types.
A stage flagged claimTask gets the kernel (not the LLM) to claim the
next ready task — or re-surface the one already claimed on a review
bounce — and inject its TaskContextAssembler bundle as L0. While a task
is claimed, plane-2's write manifest narrows to the task's affectedPaths.
Project identity is resolved from the origin-session link the task tools
already record (TaskService.projectForSession), so the loop predicate and
claim stage agree regardless of the free-form project key the agent passed.
Claiming flows through a kernel-declared TaskClaimCoordinator implemented
in apps/server (no cross-core import).
New TasksReady condition driven by EvaluationContext.readyTaskCount,
fed at resolve time via a kernel-declared ReadyTaskCounter interface
implemented in apps/server over TaskService (keeps core:kernel
decoupled from core:tasks). Loops a graph while ready>0, exits at 0.
Root Child DOX Index assembled, plus a per-module AGENTS.md across the tree
(core/*, infrastructure/*, apps/*, testing/*, and docs/examples/frontend/etc),
each following the DOX section shape: Purpose, Ownership, Local Contracts,
Work Guidance, Verification, Child DOX Index.
- editorLines now soft-wraps long logical lines at the box width (caret tracked
across wrapped rows) instead of overflowing/truncating
- alt+backspace deletes the word before the cursor (deleteWordBack)
- handle tea.PasteMsg so bracketed paste (Ctrl+V / Ctrl+Shift+V / right-click)
drops clipboard text into the focused composer; was silently dropped
- render the composer (in-session + launcher) borderless: top rule + flush text,
no side borders or trailing padding, so a terminal drag-copy yields just the
typed text instead of box chrome and whitespace
file_read on a directory/file returns newline-bearing summaries; clip kept the
newlines, so the action row rendered multi-line with a background-filled stripe
and the listing leaking underneath. Flatten the summary to one line first.
Add "tasks" and "task-detail" kinds to PreviewFrame, backed by a
sampleTaskBoard work graph that exercises every readiness glyph and
status color (epic blocked on children, a ready child, in-flight, done).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Each task row gets a 2-col glyph — ● ready (workable now) / ○ waiting (blocked) /
blank — and the detail pane shows "ready to work" or "blocked — waiting on <ids>",
fed by the new GET /tasks ready/blockedBy fields. Makes a decomposed graph legible
at a glance; a dependency-blocked task is status TODO (not the red BLOCKED lifecycle
state), so the glyph is what distinguishes them.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
GET /tasks now reports dependency-graph status per task — ready (TODO with no unmet
deps) and blockedBy (ids of unfinished blockers) — resolved via TaskGraph over each
task's FULL project board, so a status filter can't hide a blocker. Default-valued
fields are omitted (encodeDefaults=false); consumers read absent as zero.
The task_decompose approval card now shows a readable graph (epic + numbered children
with their "after" dependency labels) via renderDecomposePreview wired into
computeToolPreview, instead of the truncated raw JSON the generic fallback gave.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Block writing a file whose on-disk content changed since the session read it — a
concurrent/external edit the agent's view doesn't reflect. file_read records a
content hash on whole-file reads; the orchestrator's completion path now carries
the tool's structuredOutput (it previously dropped it, unlike SandboxedToolExecutor
— the two paths were inconsistent), so SessionContext.readHashes folds it.
StaleWriteRule compares the current hash (WorldProbe.contentHash) against the
read-time hash and BLOCKs a mismatch, telling the agent to re-read first. Files the
session wrote itself are excluded, so its own edits never look stale; partial reads
set no baseline.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Completing a task that declared affected_paths but saw no matching write this
session is now blocked ("marked done without doing it"), escapable with force.
TaskUpdateTool reads the session's writes through a SessionWrites port whose
adapter folds SessionContext (writes = recorded receipt.affectedEntities), and
matches them against the task's affected_paths globs.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claiming a task now emits SessionWorkingTaskEvent{taskId, affectedPaths} into the
session's own stream (via a SessionFactRecorder port + event-store adapter), and
SessionContextProjection folds it into SessionContext.activeTask. So a gate learns
"which task is this session working, and its scope" by reading the session stream
— no cross-stream scan, no kernel->tasks dependency. Re-emitted on an affected_paths
edit by the claimant, so the snapshot scope stays current; latest wins on replay.
This is the data source the write-scope gate needs (next).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A file_read of a path that is inside the workspace but does not exist is now
BLOCKED with "no such path" instead of silently returning empty — so an agent
that hallucinated a filename fails loud and self-corrects (list the directory,
fix the name) rather than proceeding on an absent reference. Out-of-workspace
targets stay PathContainmentRule's concern, so the gates don't double-handle.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
An LLM that edits or overwrites a file it never read is acting on hallucinated
contents. This plane-2 ToolCallRule blocks it: dispatching on FILE_WRITE (covers
file_write and file_edit), a write to a path that exists on disk but was not
file_read earlier this session is BLOCKED. A brand-new file passes — you can't
read what doesn't exist, and creating is legitimate.
- ReadFilesProjection folds the session's file_read events (request + completion,
so a failed read doesn't count) into the set of paths read, mirroring
EgressAllowlistProjection; threaded into ToolCallAssessmentInput.sessionReadPaths
via SessionOrchestrator.resolveSessionReadPaths.
- Read paths and the write target are both resolved through the probe (toRealPath),
so spelling differences and symlinks can't dodge the gate.
- The orchestrator already feeds a block's rationale back as a tool result, so the
agent reads the file and retries; the rejected-event reason now carries that
rationale too (specific audit trail instead of a generic string).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The doctrine says "search before creating", but that leans on the LLM; this is
the backstop against an agent re-creating a task across runs. TaskService.
findDuplicates matches active (non-terminal) same-title tasks, case- and
whitespace-insensitively — a DONE/CANCELLED look-alike is not a duplicate, since
that work may legitimately recur.
- task_create rejects a duplicate, naming the look-alike and offering force=true.
- POST /tasks → 409 with the duplicates, or force:true to override;
correx task create --force.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The DEPENDS_ON/BLOCKS link types were inert metadata — nothing reasoned about
them. TaskGraph turns them into answerable questions: what a task waits on
(its DEPENDS_ON targets plus any task that BLOCKS it), whether it is ready
(TODO with no unmet blocker — a terminal dependency stops blocking), and what
finishing it would unblock. Readiness surfaces workable tasks to claim; it
never assigns (honouring the rejected /tasks/next scheduler).
Surfaced across the stack:
- TaskService.ready / blockers / blocking.
- Context bundle: a `blocked` flag + `blocked_by` list, rendered as a prominent
"BLOCKED — waiting on ..." line, so an agent calling task_context learns it
should wait before charging at the task.
- GET /tasks?ready=true, GET /tasks/{id}/blockers; correx task ready / blockers.
Cross-project dependencies are scoped to a project for now (documented in
TaskGraph). Tests cover both link directions, terminal-dependency clearing,
the ready filter, the reverse blocking direction, the bundle flag, and REST.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Surfaces a task's lifecycle straight from the event log (the source of truth) so
you can see what actually happened to it — claim/submit/complete, notes, links,
git-driven status — rather than only its current state. Useful for debugging a
live agent run, where current state alone doesn't say how it got there.
- TaskHistory.render: pure timeline renderer (split content/lifecycle to stay under
the complexity cap, mirroring DefaultTaskReducer).
- TaskService.history(id): the task's own events, in order; survives soft-delete
(the deletion is part of the trail), empty for an id that never existed.
- GET /tasks/{id}/history (200 text timeline, 404 when the id never existed) and
the `correx task history <id>` CLI command.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A freestyle stage's tools are LLM-authored in the execution_plan. At run time an
unresolvable tool name is silently dropped (resolve -> null -> mapNotNull), so the
stage runs toolless — e.g. a misspelled task_update means the implementation stage
quietly loses task tracking, invisible without a live run.
ExecutionPlanCompiler now takes the registered-tool universe and rejects a plan that
names an unknown tool, with a message identifying the tool and stage. FreestyleDriver
already turns a compile failure into a rejection the architect retries. The param
defaults to empty (validation skipped) so existing callers are unaffected; Main feeds
it toolRegistry.all().
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
correx task list/search/show/create/claim/complete/export over the REST
surface, with --json passthrough and pure render helpers. Completes the
human entry points alongside the TUI board.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A disposable Markdown view of the board, grouped by status with checkboxes
and claimants — rendered from the event log, never a source of truth.
TaskMarkdown.render (pure) backs GET /tasks/export[?project=].
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
POST /tasks/sync-git advances task status from commit messages: a mention
moves a task to IN_PROGRESS, a closing keyword (fixes/closes/resolves
auth-12) walks it the legal path to DONE. Only ever advances (never
regresses or touches BLOCKED/CANCELLED), so re-running is idempotent; each
advance leaves a [git <sha>] note. Acts on commits in the request body or
reads the repo's recent log (GitCommandCommitReader) — wire it to a git
hook or CI step.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Ranked exact/substring search: TaskSearch (all terms AND, ranked title >
key > goal > criteria > notes) over one project or the whole board via
TaskService.search. Surfaced as GET /tasks?q=, a read-only task_search tool
for agents, and a `/` filter in the TUI board.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add a cross-project task board to the TUI (T / palette): a roster fetched
from GET /tasks with vim-style nav, refresh, and an enter-to-open detail
pane (goal, criteria, paths, links, notes) built from the list payload — no
extra fetch. Server: TaskService.listAll() and a project-optional GET /tasks
(scoped with ?project=, whole board without; recent-first).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The context bundle resolved only TASK and DOC link targets; ARTIFACT and
SESSION stayed raw. Add two ports (TaskArtifactResolver/TaskSessionResolver)
with apps/server adapters: artifacts resolve to producing stage/session +
content excerpt from CAS, sessions to status/intent/workflow. Unresolved
targets still fall back to raw links. Wired through the tool and REST bundle.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
CRUD, lifecycle transitions, work-graph links and notes over /tasks;
share link-kind inference (TaskTargetKinds) between the REST route and
task_update tool. End-to-end route tests.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds task tracking as a first-class event-sourced aggregate (ADR-0012). The
event log stays the only source of truth; the board is a projection, and all of
a project's task events live in one stream (tasks:<projectId>).
core:
- new core:tasks module mirroring core:sessions: TaskStatus/State/Reducer/
BoardProjector/CounterProjection/Repository + TaskService write path that
appends events via the existing EventStore (no new storage)
- task events in core:events (sealed marker TaskEvent), registered in
Serialization.kt; status is derived by the reducer from lifecycle facts
- work-graph edges: TaskLinkType + typed TaskTargetKind on the link, so bundle
resolution dispatches on the recorded kind instead of guessing from the id
- delete is a soft tombstone (TaskDeletedEvent); cancel stays a visible outcome
surfaces:
- agent tools (Tier T2 mutators, T1 read): task_create / task_update /
task_delete / task_context, registered in both the main and per-workspace
tool registries
- REST GET /tasks/{id}/context for external agents
context bundle (TaskContextAssembler):
- task fields + acceptance criteria + relevant files + resolved dependency
tasks (live status) + raw related links + notes, with a compact render()
- semantic enrichment via a TaskKnowledgeRetriever port bridged to the existing
L3 retriever (late-bound holder for composition-root ordering)
- ADR/doc resolution via a TaskDocumentResolver port + filesystem adapter
(docs/decisions/adr-*.md, padding-agnostic; *.md paths)
Unit-verified across core:tasks / infrastructure:tools / apps:server; full
gradlew check green. Not yet live-QA'd end-to-end against a running server.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
renderStats used the default-locale String.format, so under a comma-decimal
locale (e.g. ru_RU) percentages/rates rendered as "58,3" — both producing
locale-dependent CLI output and failing StatsRenderTest. Format all five
numeric rows with Locale.ROOT.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Two v2-unlocked polish items now that the TUI is on bubbletea v2:
Native cursor — the composer caret was a frame-animated "▏" that forced the
redraw loop to keep ticking in insert mode. v2's View.Cursor lets the terminal
draw and blink a real cursor, so:
- editorLines/launcherInput drop a width-1 marker rune (U+E123, same cell width
as "▏") at the caret; View() locates it in the fully-composited frame via
splitCursor → (X,Y), swaps it for a space, and sets a blinking CursorBar there.
- render() (preview/golden tests) resolves the marker to the static "▏" glyph, so
screenshots are unchanged and the marker never leaks into output.
- nil cursor in normal mode hides it (vim-correct); gated on overlay==None so an
open palette/files modal (which keeps editMode==Insert) doesn't draw a cursor
behind it.
- animating() no longer forces a redraw in insert mode — the terminal blinks the
cursor itself, so an idle insert screen holds still (native text selection
survives) and burns no frames.
Window title — View.WindowTitle names the terminal tab after the active session,
falling back to the model name then "correx".
Tests: new cursor_test.go (coord extraction, mode/overlay gating, no marker leak);
updated the animating-gate test for the new insert-mode contract. Build, vet,
gofmt, full suite green; compose preview shows the static caret with no marker leak.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Move the Go TUI from github.com/charmbracelet/{bubbletea,lipgloss} to the v2
ecosystem at charm.land/{bubbletea,lipgloss}/v2 (Go 1.25). v2's kitty keyboard
disambiguation is on by default, so Shift+Enter now reliably inserts a newline
in the composer (Ctrl+J / Alt+Enter kept as fallbacks for non-kitty terminals).
Approach: rather than rewrite ~190 key-match sites to v2's (Code, Mod) idiom, a
small shim (internal/app/key.go) converts a v2 KeyPressMsg into the v1-shaped
keyMsg the handlers already expect, at the single Update boundary. The rest is
mechanical:
- key constants tea.Key* → shim consts; tea.KeyMsg → keyMsg.
- lipgloss.Color is now a func returning color.Color, not a type → fields/params
retyped to image/color.Color (theme/diff/overlays/view).
- v2 View() returns tea.View: render() builds the string, View() wraps it and
carries AltScreen (alt-screen is a per-frame View field now, not a program opt).
- WithWhitespaceBackground/Foreground → WithWhitespaceStyle(Style).
- preview: drop lipgloss.SetColorProfile (v2 renders truecolor by default).
Build, vet, gofmt, and the full test suite are green; preview renders truecolor
across kinds.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Press d (or the "panel" palette command) in-session to cycle the right panel:
- events — the live event stream (default, unchanged)
- changes — a token-usage line (tokens + turns, summed from the transcript's
TurnMetrics) over a git-status-style list of files the session has
written, each with summed +/− from its diff. ^x still opens the full
diff.
- off — hide the panel; the output transcript takes the full width.
changesRows derives everything from data already in the model (router-turn
metrics + the tool entries' unified diffs), so no new protocol. Footer + ? help
updated; "changes" preview kind added.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Enter still submits; Ctrl+J (always) and Alt+Enter (most terminals) insert a
newline at the cursor. True Shift+Enter isn't distinguishable from Enter in
bubbletea v1.3.10 (terminals send the same byte, no kitty-protocol parsing), so
these are the working stand-ins — noted in the input hint.
- editorLines() renders the buffer as styled rows with the caret at the cursor,
splitting on \n and capping at maxInputLines (last rows kept). Both the idle
launcher input and the in-session bottom bar use it and grow to fit; View()
sizes the bottom bar via inputBarHeight() instead of the fixed inputH.
- Footer gains a "^J newline" hint in insert mode (non-filter); the launcher
sub-line shows "⌥↵/^J newline". New "compose" preview kind + editor_lines_test
(split + height cap).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The launcher removed the idle session list, so the per-row date moves to the R
resume overlay — now the sole session list. sessionListRow shows an absolute
local date (absDateISO → shortDateTime) alongside the relative age:
"Jun 22 14:30 · 5m ago", with the year shown for older-than-this-year rows.
Delete the now-unused idle-panel renderers (sessionRows / welcomeRows /
workflowRows) and renderMainNarrow's dead idle branch (renderLauncher owns idle,
narrow included). Add a "resume" preview kind to screenshot the R overlay.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replace the idle two-panel layout (session list + welcome, where the welcome
panel duplicated the list) with an opencode-style launcher: a compact, centered
input whose lower-right shows the launch target and model (chat by default), a
hideable right rail of status + quick keys, and no session list — it's a keypress
away via R.
- Tab (or w) cycles the launch target: chat → workflows → chat (launcherWf),
shown at the input. Enter on chat starts a chat; on a workflow, StartSession
with the typed text as the brief.
- Right rail (connection, session/active counts, i/Tab/R/?/p keys) hides via the
new "rail" palette command. Drops automatically on narrow terminals.
- View() suppresses the bottom input bar on idle (the input is the launcher);
footer + ? help updated to the launcher keys (help-coverage test extended).
The old idle-panel renderers (sessionRows/welcomeRows/workflowRows) are now
unused but kept in the tree pending Phase 2 + a decision on where the session
dates should live now that the idle list is gone. Multi-line input (Ctrl+J /
Alt+Enter newline) is Phase 2.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Was the UTC-then-local time-of-day; now the same "Jan 02 15:04" / "Jan 02 2006"
shortDateTime as the session list, so a previous-day session isn't shown as if
it were today.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
formatTime forced .UTC(), so every event-log timestamp and the welcome "recent"
times read hours off the operator's wall clock — and inconsistent with the local
date just added to the session list. time.UnixMilli is already local, so the
explicit .UTC() was the bug; drop it. No test asserts a clock string (the render
matrix only checks event type/detail), so this is display-only.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The collapsed tool row showed "tool output (N chars)" where N was len() of the
raw unified diff — i.e. the diff blob's *byte* length (headers, @@, +/-/context
all counted), labelled "chars". That number corresponded to nothing the operator
cares about, and len() is bytes not characters.
- Summarise a write/edit by the diff's row count instead — "· diff (7 rows) —
^x to view" — matching exactly what the ^x modal reports. Non-diff output (the
defensive fallback) now counts runes, not bytes, so "N chars" is truthful.
Retire itoaLen (the byte-count-as-"len" footgun).
- Harden the (+a −b) line counts in diffSummary: guard the file header with the
trailing space ("+++ "/"--- "), so a changed line whose content starts with
"++"/"--" is counted instead of mistaken for a header. Apply the same guard in
parseUnifiedDiff so such a line is rendered (and counted) rather than dropped.
tool_summary_test.go covers the diff-row summary, the rune (not byte) count, and
the ++/-- content-line edge case.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Audited handleNormalKey / handleApprovalKey / the OverlayDiff handler against the
help cheat-sheet and found gaps: enter, ↑↓/jk, /, s, c, a, w, q and ? itself were
bound but undocumented. Reorganise helpBody into navigate / session / overlays /
approval band / diff-viewer groups covering all of them. help_complete_test.go
asserts a row exists for every binding so the help can't silently drift from the
handlers. The longer sheet scrolls fine via the modal-scroll support.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The idle "sessions" panel listed id/status/name/stage but no time at all. Add a
right-aligned last-activity timestamp (Session.LastEventAt) per row — "Jan 02
15:04" within the current year, "Jan 02 2006" for older — so old sessions are
distinguishable at a glance. The name is truncated first so the box's width
clamp never eats the date column. (The R resume overlay already shows a relative
age; this is the live list on the starting screen.)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The EVENTS panel showed "● live" off m.connected alone, so a completed/failed/
paused session kept reading "live" as long as the websocket was up — stale, since
a terminal session emits no more events. Derive the badge from the selected
session's Status (the same vocabulary the status bar uses): ✓ done, ✗ failed,
‖ paused, ● live while active, ○ idle when disconnected. event_badge_test.go
covers all five.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Both were scaffolding placeholders from Epic 14 (2c459da) that were never
built out: each was just an `application` build.gradle (clikt only, no
internal deps) and a one-line `println` Main.kt. Nothing depended on them —
no project(...) reference anywhere, only the settings.gradle includes, and no
script/doc/CI launched their MainKt — so they only cost build time configuring
two inert modules.
Delete both module dirs and drop their include lines from settings.gradle.
`./gradlew projects` now lists only :apps:cli and :apps:server (configures
clean).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The ? help and S stats modals rendered at fixed height; compositeOverlay
drops any rows past the screen, so on a 24-30 row terminal their bottom
sections — including the close hint — vanished with no way to reach them
(and the backdrop border corrupted). Recent help additions worsened it.
- Add a shared scrollable-modal renderer: helpBody()/statsBody() produce
discrete lines, renderScrollModal() windows them by m.modalScroll (clamped
to the viewport via modalBodyRows/scrollableModalMax) and pins the hint, so
the box never exceeds the screen. A "(off-end/total)" indicator shows in the
title when scrolled. Keys: ↑↓/jk line, PgUp/PgDn page, ^u/^d half, g/G ends;
any other key still dismisses help. modal_scroll_test.go (2): clamp + no-op.
- Page the artifacts viewer (v) too: PgUp/PgDn were one line at a time — now a
full body, with ^u/^d half and g/G ends via scrollArtifact().
Verified via cmd/preview: help/stats close hints now survive at h=20-28 (were
clipped below ~32/40) and show the scroll indicator; full at tall heights.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The ^x fullscreen viewer (OverlayDiff) backs every tool preview — split diff
for write/edit, the command card for shell, plain lines otherwise, plus any
transcript tool-output — but only scrolled one line per keypress, making a
long diff tedious to read.
- Add paging to the viewer, matching the OUTPUT free-scroll contract: ↑↓/jk
by line, PgUp/PgDn by a full body, ^u/^d by half, g/G to the ends. New
scrollDiff() clamps to [0, diffMaxScroll] so paging past either edge lands
exactly (no dead presses). diff_scroll_test.go (2): clamp + no-op-when-fits.
- Update the modal hint to advertise paging (line / page / ends / close) and
add a "diff / preview viewer (^x)" section to the ? help overlay. Also note
the T3+ confirm-again behaviour in the approval-band help.
Approval ergonomics re-audit (BACKLOG §E): the spec'd band still holds —
single-key y/n/e for T0-T2, mandatory second-key confirm for T3+ (HighTier),
navigable queue with an i/n indicator, docked band never modal-stacked; all
covered green by approval_test.go. The scroll granularity above was the only
gap, now closed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Backlog hygiene per the file's own maintenance rule — resolved items get
moved out, not left ticked-in-place.
- Cut resolved body bullets from BACKLOG §A/B/C/D/E/G/I (health probes,
static-first infra, CritiqueFinding, architect contradiction-check, A1/A2
gates, batch fetch-approval, source/egress events, render matrix, markdown,
session list, idea promotion, turnId/probe/narration, .cs symbols, scene
validator). All already archived in RETRO's burndown table.
- Narrow partially-done items to their live follow-up rather than delete:
B5 -> static_check stage emitter only; B6/C-A3 -> CritiqueOutcomeCorrelated
producer only; C-A1 -> planner-prompt activation only.
- Drop the stale "§E UNBLOCKED" index line (plan-comparison is BLOCKED, no
PlanCandidate event; idea-promotion shipped f107ff5).
- Archive the operator-requested TUI wave (palette, @ picker, CLAUDE.md L0
injection, inline action rows, free-scroll/help/filter) in RETRO 2026-06-22
— never BACKLOG items, would otherwise be orphaned in neither file.
- gitignore apps/tui-go/preview (cmd/preview build artifact).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Three TUI-polish items.
- OUTPUT free-scroll: PgUp/PgDn (and ^u/^d half-page) scroll the transcript
back through history; esc snaps to tail-follow. outputScroll is clamped
against outputViewport(), which mirrors View/renderMain geometry, so the
edges land exactly (no dead presses unwinding an overshoot). routerRows
split into buildTranscriptRows (full render) + windowing so the handler
measures the same content. Tests cover the clamp + the fits-in-panel no-op.
- Help overlay: `?` (or the "help" palette command) opens a keybinding
cheat-sheet grouped by context (session / overlays / approval band) — the
non-palette keys especially. Any key dismisses it.
- Event-inspector filter: `/` filters the event list by a case-insensitive
substring of type/detail (currentEvents applies it); `c` clears, esc stops
editing then closes. Fresh per open. EVENTS side panel stays unfiltered.
go build / vet / test green; rendered help + filtered inspector via
cmd/preview (help, events-filter kinds).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The ⏵ tool-start row doubled up with the ✓/✎ result row in OUTPUT. Drop it
inline — tool starts still show in the tool list + EVENTS panel. The inline
transcript now carries only outcomes: ✎ writes, ✓/✗ tool results, ⌘/✕
approvals, ⊞/⊟ grants.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Surface side-effecting ("external feedback") events in the conversation
flow, opencode-style, instead of only in the EVENTS side log.
- New RouterEntry role "action" (glyph + text) appended at arrival so it
interleaves with chat turns by order. Rendered dim with an accent gutter
glyph; the EVENTS panel still carries the full stream.
- Surfaced: tool start (⏵), tool done (✓ / ✎ wrote <path> (+a −b) for a
diff), tool failed (✗), rejected (✕ blocked), approval resolved
(⌘ approved / ✕ rejected, noting "· via grant" on a grant auto-approve),
and grant create/revoke (⊞ / ⊟ · scope). ApprovalResolved names no tool
on the wire, so the tool is recovered from the pending queue before the
gate is dropped.
- Toggle: "inline actions" palette command flips actionsHidden (rows are
always recorded, gated at render, so toggling reveals history); persisted
in tui-prefs.json. prefs.go refactored to load/mutate/save so the new
field and statusbarHidden no longer clobber each other.
- demo.go: "actions" preview kind.
go build / vet / test green; rendered the flow via cmd/preview.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Surface the new cross-session grant scopes in the TUI.
- Approve-always (A) now opens a scope picker instead of immediately
creating a SESSION grant: choose this session / this project / everywhere.
SESSION stays the default (A then enter/s = old behaviour); p and g
create PROJECT / GLOBAL grants. The grant + approval are sent on confirm
(esc cancels, leaving the call pending).
- New "grants" palette command + G shortcut open a standing-grants viewer
(OverlayGrants) listing the active PROJECT/GLOBAL grants — scope, tool,
permitted tiers, project path — with x/enter to revoke. The server's
RevokeGrant reply (a fresh grant.list) refreshes it in place.
- protocol.go: TypeGrantList + GrantDto; RevokeGrant/ListGrants encoders;
CreateGrant now documents the wider scopes. server.go decodes grant.list
into m.grants and treats it as a non-session global reply.
- render-matrix coverage: grant.list classified as a non-rendering query
reply (populates the overlay, not the transcript).
- demo.go: grants + grant-scope preview kinds.
go build / vet / test green; rendered both overlays via cmd/preview.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
kami