fix(ui): correct markdown cache identity, streaming churn, and redundant tiers

Review follow-up on the #2769 highlight caches.

Fingerprint strength. The block/highlight caches are now global and
content-addressed, so a hash collision no longer mis-colors a block — it returns
a *different* block's rendered HTML and shows the user source they never wrote.
Length + one 32-bit FNV-1a is not enough key space for that failure mode at
session scale. `contentFingerprint` now combines two independent 32-bit
multiplicative hashes with a final avalanche (~64 bits); two multiplies per
character are free next to Shiki tokenization.

Streaming churn. Content addressing made every streaming step of the trailing
`live` block insert a new cache entry, so one long message evicted the settled
`full` blocks the fix exists to keep warm. `full` and `live` blocks now use
separate caches; the live cache is small (32 entries / 2MB) because it only has
to absorb repeat renders of the same step.

Redundant worker-side caches. `markdown-worker.ts` is the only sender to the
Shiki worker, and its client cache is larger than the worker-side ones, so the
worker caches could not serve a hit the client had not already served — they
only duplicated up to 48MB of payloads in a second heap. Removed; the reason
memoization belongs on the client is now documented there, along with why only
`highlightTokens` carries a theme in its key.

Dead `cacheKey` plumbing. `renderMarkdownBlocks` kept a `cacheKey` parameter it
only `void`-ed. Removed it and the now-unused `useMorphdomMarkdown` prop; the
remaining call-site local is renamed `fadeKey` for what it actually keys.

Tests: image-mode cache identity, streaming-does-not-evict-settled-blocks,
live-cache reuse, and a 20k same-length-source fingerprint collision check.
Each new guard was verified to fail without its fix.

CHANGELOG.md

@@ -5,6 +5,7 @@ All notable changes to this project will be documented in this file.
 ## [Unreleased]
 
 - **Settings/Integrations:** a new Integrations settings page lists Claude Code, Command Code, and Cursor plugins with install, update, setup, and remove actions, plus Discord and Telegram Coming soon placeholders.
+- **Chat:** an open conversation no longer keeps re-coloring the same code blocks in the background, so browsing files with a chat open stops pinning a CPU core and spinning up the fans (thanks to @makeittech).
 - Usage/Claude: Claude plan limits now work when you are signed in through Claude Code, without also signing into Anthropic in OpenCode; the account is read from Claude Code's own login on macOS, Linux, and WSL. The page shows your session and weekly limits again, adds per-model weekly limits and extra usage spending, and names your plan. Limits are kept on screen instead of disappearing when Anthropic temporarily blocks refreshes.
 - Git: the pull request panel now follows the branch's current open PR, and an open PR always wins over an older merged or closed one. After a PR is merged or closed the panel keeps showing it as the branch's last PR and offers creating the next one right below it (thanks to @makeittech).
 - 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.