fix(ui): stop the context meter from counting every internal round-trip

The token breakdown of an assistant message accumulates across every API
round-trip inside the turn: each tool call re-reads the whole cached
prompt, so input/cache.read add up to several times the context window.
Every context-usage surface summed those fields, which is why the meter
could read 330% of a 1M window whose real fill was 232,872 tokens
(23.3%), and why reopening an older session jumps the readout (#2562).

The server reports the final round-trip's window as tokens.total
(optional in the message schema; opencode 1.18.18 returns it, verified
against its live /session/:id/message API). Prefer it everywhere the
window fill is displayed and fall back to summing only when the server
did not send it: contextTokensFromBreakdown in tokenUtils now owns that
rule, and the context store extractor, sync store getter, work status
panel, context sidebar, VS Code layout, mini chat, and mobile metadata
all use it instead of their own inline sums.

Fixes #2562

packages/vscode/CHANGELOG.md

@@ -2,6 +2,7 @@
 
 - Chat: new chats no longer start against a deleted last worktree directory; they fall back to the active project instead of saving the first message and never starting.
 - Chat: opening a busy subagent in the context panel now shows its history instead of only the working-status line (thanks to @makeittech).
+- The context usage readout no longer climbs over 100% after turns with many tool calls and no longer jumps when reopening an older session; it now shows what the window actually holds.
 - Projects: project names now match the folder name exactly, so `.ssh` and `opencode-claude` are no longer shown as `.Ssh` and `Opencode Claude`; names you renamed yourself are kept.
 - Skills Catalog: the source is now named ClawHub instead of "ClawdHub" (thanks to @makeittech).