Perf: cut worker-thread saturation/CPU from session diffs, snapshots, SQLite contention, and event fanout (PR #48638 series)

Context

Part of the fork hardening/perf series proposed in #48638 (integration branch session-diff-events, fork AceRothstein71/opencode). Original tracker: #49565.

PR #48638 is intended to close this issue.

Problem

Under concurrent sessions / parallel agents: worker-thread saturation from permission matching, turn-diff recomputation, event fanout to TUI/server clients, snapshot churn and snapshot-index recovery, and SQLite contention across sessions. Symptoms: stalled turns, high CPU, delayed diffs.

Change set (commit list, session-diff-events)

• 41b33fee30 processor coalesce
• 562dccf3c0 processor coalesce cleanup
• fdb3778f95 audit findings: diff/events/sqlite
• 857ce25bb8 doom-loop reload
• c1fa9435ec wave 1-5 reliability
• 15315be778 wave 6 residual
• 56feb5c84e wave 7-11 hardening
• 59c3e3280f sqlite lanes
• 3d64416ba6 tui bridge sync
• d84bbc617b session event-loop load
• 6c1ddb3e4a snapshot heal
• d27f83c334 snapshot index recovery
• a38c279be3 vcs diff context
• 2d1e6da481 core directory index
• dc7e2ad448 wave G

Validation

• Owner-observed: CPU usage is definitely lower under identical parallel-agent workloads.
• Worker-thread stalls sharply reduced; turn-diff path bounded; mock-verified tool schema (206 tools, zero invalid required, zero empty combinators).

Linked: PR #48638 · tracker #49565 · sibling tracking issues listed in a follow-up comment.

Start with PR #48638 and the session-diff-events integration branch, then review the listed commits covering worker processing, event fanout, snapshots, VCS diffs, and SQLite lanes. Done means reproducing the parallel-agent workload and confirming lower CPU, fewer worker stalls, bounded turn-diff work, and the stated schema validation results.