feat(history): CR-5 windowed reads — ?limit= / ?beforeSeq= on GET /conversations/:id

Selection sinceSeq < seq < beforeSeq; newest-limit window, ascending; positive-
integer validation (400, store never sees an invalid window); 1-based gap-free
seq codified as the contractual has-older mechanism (no earliestSeq field).
transport-contract 0.9.0->0.10.0, wire 0.6.0->0.6.1 (doc-only).
conversation-store +8 tests, transport-http +20; 935 vitest + 112 bun green.
Live-verified: 6/6 probe checks OK. FE courier: frontend-history-windowing-handoff.md

1 files changed, 74 insertions, 6 deletions

diff --git a/tasks.md b/tasks.md
index f77577a..205d9c9 100644
--- a/tasks.md
+++ b/tasks.md
@@ -26,8 +26,11 @@ extension through the host):
 
 ## How to run
 ```bash
-KEY1=$(grep DISPATCH_API_KEY_OPENCODE1 .env | cut -d= -f2)
-PORT=4567 DISPATCH_API_KEY="$KEY1" bun packages/host-bin/src/main.ts   # boots app + collector
+# .env auto-loads DISPATCH_API_KEY (do NOT re-export) and pins BACKEND_PORT (beats PORT).
+# Private probe instance: override the port + ISOLATE data paths (ORCHESTRATOR §8):
+BACKEND_PORT=4567 SURFACE_WS_PORT=4569 DISPATCH_DB=/tmp/opencode/probe/dispatch.db \
+  DISPATCH_TRACE_DB=/tmp/opencode/probe/traces.db DISPATCH_JOURNAL=/tmp/opencode/probe/app.ndjson \
+  bun packages/host-bin/src/main.ts   # boots app + collector
 curl -s -X POST localhost:4567/chat -H 'content-type: application/json' \
   -d '{"conversationId":"c1","message":"Say hello in 3 words."}'        # field = conversationId
 ```
@@ -323,8 +326,35 @@ FE courier in: `../dispatch-web/backend-handoff-cache-warming.md` (+ CR-1/CR-2 f
   2 automatic warms @5s each pushing future `nextWarmAt`, mid-turn close → `abortedTurn:true` +
   `done.reason:"aborted"` + warming disabled, catalog scopes + table field present, echo present.
 
+## History windowing — FE CR-5 (DONE)
+FE courier in: `../dispatch-web/backend-handoff-chat-limit.md` (+ living `backend-handoff.md` §2
+CR-5). Courier out: `frontend-history-windowing-handoff.md`. User-gated call: ask #3 shipped as
+the INVARIANT option (no new field) — seq is contractually **1-based, monotonic, gap-free**; FE
+derives `hasOlder` from `chunks[0].seq > 1`.
+- **Wave 0 (orchestrator, contracts):** `limit`/`beforeSeq` query-param semantics + validation +
+  `latestSeq` windowed-read caveat documented on `ConversationHistoryResponse`
+  (`@dispatch/transport-contract` `0.9.0→0.10.0`); 1-based seq guarantee codified on
+  `StoredChunk` (`@dispatch/wire` `0.6.0→0.6.1`, doc-only).
+- **Wave 1 — `conversation-store`:** additive `loadSince(id, sinceSeq?, window?: { beforeSeq?,
+  limit? })` — selection `sinceSeq < seq < beforeSeq`, newest-`limit` window, result stays
+  ascending; garbage-in treated as absent (transport validates upstream). +8 tests.
+- **Wave 2 — `transport-http`:** parses + validates the params (positive integers; malformed/
+  zero/negative → 400 `{ error }`, store never called with an invalid window); two-arg call
+  shape preserved when no params (regression-guarded). +20 tests.
+- 935 vitest + 112 bun tests, typecheck + biome clean. **LIVE-VERIFIED** (isolated boot, real
+  flash turns): firstSeq=1; `limit=2`→`[5,6]` ascending w/ correct `latestSeq`; `limit=9999`→
+  full log; `beforeSeq=3`→`[1,2]`; `beforeSeq=3&limit=1`→`[2]`; `limit=0`/`beforeSeq=0`/
+  `limit=abc`→400×3. `RESULT: OK` ×6.
+- **Scar tissue (process):** (1) probing with a PRIVATE boot was overkill — the windowing checks
+  are read-only GETs and the dev stack was running; prefer probing `bin/up`/`up2` or asking the
+  user (ORCHESTRATOR §8 updated). (2) The §8 boot recipe was stale (`DISPATCH_API_KEY_OPENCODE1`
+  doesn't exist; an empty re-export OVERRIDES `.env` → "No providers registered"; `.env`'s
+  `BACKEND_PORT` beats `PORT`; un-isolated data paths spawn a duplicate collector on the dev
+  DB) — recipe fixed in §8 + above. (3) Violated the bracket trick once (`pkill -f 'cr5-data'`
+  self-matched → killed parent shell, timeout-with-no-output); the existing §8 rule stands.
+
 ## Open items
-- **Context window LIMIT (next, sibling of context size):** expose the selected model's max
+- **Context window LIMIT (deferred, sibling of context size):** expose the selected model's max
   context-window token limit so the FE can render `contextSize / limit` (e.g. `1286 / 200000`).
   Source = the provider/model catalog (`ModelInfo`); likely a field on the models response
   and/or stamped alongside `contextSize`. "context size" = current usage (DONE); "context
@@ -341,9 +371,47 @@ FE courier in: `../dispatch-web/backend-handoff-cache-warming.md` (+ CR-1/CR-2 f
   chain).
 
 ## Roadmap
-1. **CLI** — done.
-2. **Web frontend** (in progress, SEPARATE repo `../dispatch-web`; Svelte +
+1. **Web frontend** (in progress, SEPARATE repo `../dispatch-web`; Svelte +
    DaisyUI, same methodology). Slice 2 = browser chat MVP consuming the
    wire/transport-contract + metrics. Cross-repo contract changes are couriered
    via the user (ORCHESTRATOR §7); `lsp references` does not span repos.
-3. **dedup / storage growth** — after the frontend (see Open items).
+2. **CLI → open-tab handoff (cross-client messaging):** an AI given a short
+   conversation identifier (4+ chars) in chat can use the CLI to send a message
+   INTO that conversation, and an FE tab watching it sees the message + reply
+   live. To the receiving AI the message appears as a REGULAR user message
+   (no special framing). The broadcast plumbing already exists and is
+   live-verified (detached turns + `chat.subscribe` + CR-3 `user-message`
+   event); the gaps are (a) a conversation LIST/discovery endpoint (none
+   exists), (b) short-id / prefix resolution of `conversationId` in the CLI,
+   (c) end-to-end verification of the handoff workflow. CLI semantics:
+   - **read last message** (new): BLOCKING — waits until the in-flight turn
+     settles, then returns ONLY the AI's last message of that turn (not the
+     whole conversation).
+   - **send, no `--queue` flag (default):** BLOCKING — sends, waits for the
+     turn to settle, returns the AI's last message (same shape as the read).
+   - **send with `--queue`:** enqueues the message into the conversation's
+     message queue (roadmap item 3) and exits immediately.
+3. **Message queue + steering injection (backend core; prerequisite for the
+   `--queue` flag in item 2):** a per-conversation queue a client (FE or CLI)
+   can push a message onto while a turn is GENERATING. Delivery semantics:
+   - On the turn's next TOOL CALL, queued messages are injected as "steering"
+     messages returned alongside the tool result (the model sees them
+     mid-turn and can adjust course).
+   - If the turn ends before any tool call fires, the queued messages are
+     COMBINED and sent as a fresh user message starting a NEW turn.
+   - FE queueing UX (queue while generating) couriered to `../dispatch-web`.
+   Touches the kernel turn loop (injection at the tool-result boundary) — a
+   contract-first design pass needed before summoning.
+4. **CLI flag to open/activate an FE tab:** optional CLI flag (new or existing
+   conversation) that makes an already-open frontend open the conversation as a
+   tab and mark it active. Does NOT exist today — no backend→FE "open/focus
+   conversation" push (only the inverse, `POST /conversations/:id/close`).
+   Needs a new broadcast WS op + endpoint/flag here, and FE handling couriered
+   to `../dispatch-web`.
+5. **`todo` tool** — a per-conversation task-list tool the model maintains
+   (like opencode's todowrite/todoread), as a standard tool extension; likely a
+   surface so the FE can render the live list.
+6. **`web_search` tool** — a web search tool (like old dispatch's;
+   reference-only source at `../dispatch-source`), as a standard tool extension.
+
+(Done and dropped from the list: CLI; dedup / storage growth.)