MCP `search` and `issues` draw ui:// cards: each hit or issue one row, linked to its page #11207

Intent: the two most-read tools without a card get one, so Claude, ChatGPT and VS Code show them the way the page does.

Done:

• [ ] [ ] search draws ui://forge/search: one row per hit with its kind, title and link, the query on top
• [ ] [ ] issues draws ui://forge/issues: one row per issue with state, pillar and what it waits on
• [x] both are in CARD_KINDS and CARD_OF_TOOL (src/web/studio/cards.ts) and in docs/MCP.md's card table (its test fails otherwise)
• [x] untrusted titles keep their untrusted mark on the card; every field escaped
• [x] bun test test/mcp-size.test.ts passes without raising a ceiling

Files: src/web/studio/cards.ts, src/http/mcp/tools/search.ts, src/http/mcp/tools/issues.ts, docs/MCP.md, tests.

Reviewed pull #11215 at its head: ticked both are in CARD_KINDS and CARD_OF_TOOL (src/web/studio/cards.ts) and in docs/MCP.md's card table (its test fails otherwise); untrusted titles keep their untrusted mark on the card; every field escaped; bun test test/mcp-size.test.ts passes without raising a ceiling.

Left open after pull #11215 landed: 2 unticked items remain ([ ] search draws ui://forge/search: one row per hit with its kind, title and link, the query on top; [ ] issues draws ui://forge/issues: one row per issue with state, pillar and what it waits on). The planner files each as its own issue and closes this one; a person may close it, tick what is done, or say what is left. Until then the crew does not pick it again.

Both open boxes are shipped on main by 44d4ed381 (pull #11215): search draws ui://forge/search (cards.ts kind search, one row per hit, query on top) and issues draws ui://forge/issues (state, pillar, what it waits on), covered by test/mcp-search-issues-card.test.ts and test/mcp-issues-card-tool.test.ts. The doubled [ ] [ ] boxes kept the reviewer from ticking them. Nothing is left, so I am closing it with no new issue.