…elegate return modes

- Add Delegate Return Modes section (| silent, | silent-wake, normal)
- Add Canary Validation section with blind testing methodology
- Document secret-world pattern, 12-test matrix with results
- Document 3 failure modes: dispatch cancelled, shard confabulation, narrated dispatch
- Document confabulation finding: agents cannot self-verify enrichment provenance
- Document chain hop architecture (parent-orchestrated only, sub-agents can't dispatch)
- Document | silent-wake gap with production timestamps
- Update test count to 137 (add 8 silent-announce tests)
- Update status to v4

Addresses #188 (testing methodology) and #189 (silent-wake design).

Parse '| silent-wake' suffix on CONTINUE_DELEGATE tokens. When a silent-wake
delegate completes, the enrichment is delivered as a system event (no channel
echo) AND the parent session is woken via requestHeartbeatNow — enabling
autonomous cognition loops where the agent acts on shard returns without
external nudge.

Three delegate return modes now complete:
- Normal (default): echo to channel + wake parent
- | silent: no echo, no wake (passive enrichment)
- | silent-wake: no echo, but wake parent (autonomous chains)

Changes:
- tokens.ts: parse | silent-wake before | silent (avoid partial match)
- agent-runner.ts: thread silentWake -> silentAnnounce + wakeOnReturn
- subagent-spawn.ts: add wakeOnReturn param
- subagent-registry.types.ts: add wakeOnReturn to SubagentRunRecord
- subagent-registry.ts: thread wakeOnReturn through register + announce
- subagent-announce.ts: import requestHeartbeatNow, call on silent-wake
- tokens.test.ts: 8 new tests (parsing, case-insensitivity, hyphen/space,
  mutual exclusion with | silent)

68/68 token tests pass. TypeScript clean (0 errors in src/).

Closes #189

Parse '| silent-wake' suffix on CONTINUE_DELEGATE tokens. When a silent-wake
delegate completes, the enrichment is delivered as a system event (no channel
echo) AND the parent session is woken via requestHeartbeatNow — enabling
autonomous cognition loops where the agent acts on shard returns without
external nudge.

Three delegate return modes now complete:
- Normal (default): echo to channel + wake parent
- | silent: no echo, no wake (passive enrichment)
- | silent-wake: no echo, but wake parent (autonomous chains)

Changes:
- tokens.ts: parse | silent-wake before | silent (avoid partial match)
- agent-runner.ts: thread silentWake -> silentAnnounce + wakeOnReturn
- subagent-spawn.ts: add wakeOnReturn param
- subagent-registry.types.ts: add wakeOnReturn to SubagentRunRecord
- subagent-registry.ts: thread wakeOnReturn through register + announce
- subagent-announce.ts: import requestHeartbeatNow, call on silent-wake
- tokens.test.ts: 8 new tests (parsing, case-insensitivity, hyphen/space,
  mutual exclusion with | silent)

68/68 token tests pass. TypeScript clean (0 errors in src/).

Closes #189

…de (#189)

…nditional

One diagnostic line to trace why requestHeartbeatNow doesn't fire
for timed delegates post-reset. Logs the three values that gate the
wake path: wakeOnReturn, targetRequesterSessionKey, silentAnnounce.

Ref: #189 follow-up

The merged debug line used bare 'log.info()' which doesn't exist in
subagent-announce.ts. This file uses defaultRuntime.log(). Build was
failing on the dts step.

Ref: #189

* Initial plan

* fix(memory): tighten embedding manager inheritance types

移植上游 commit 13ae1ae。

在 embedChunksWithOpenAiBatch 和 embedChunksWithGeminiBatch 方法中添加
显式类型断言，确保 batchResult 在运行时类型检查后被正确识别为
Map<string, number[]> 类型。

这修复了 TypeScript 类型推断的不足，使得嵌入管理器的类型继承更加严格和安全。

同时移除了未使用的类型导入（MemoryEmbeddingProbeResult 和 MemorySyncProgressUpdateType），
以通过代码检查。

关联上游 commit:
- 13ae1ae: fix(memory): tighten embedding manager inheritance types
- 涉及文件: src/memory/manager.ts (原为 manager-embedding-ops.ts 和 manager-sync-ops.ts)

Co-authored-by: jiulingyun <126459548+jiulingyun@users.noreply.github.com>

---------

Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: jiulingyun <126459548+jiulingyun@users.noreply.github.com>

…penclaw#189)

…upstream

22 comment-edits across 13 files. All bare `#N` references that point
to karmaterminal/openclaw or karmaterminal/openclaw-bootstrap fork
issues now use the prefixed `karmaterminal/openclaw#N` (or
`karmaterminal/openclaw-bootstrap#N`) form so they cannot be silently
auto-linked to unrelated upstream openclaw/openclaw issues.

Citations changed:
- bug openclaw#639 -> karmaterminal/openclaw#639 (8 sites: compact-reasons.{ts,test.ts}, followup-runner.ts ×3, agent-runner-execution.ts ×3, request-compaction-tool.ts ×2)
- (#580) / See #580 -> (#580) / See #580 (4 sites: context-pressure.{ts,test.ts})
- (#531) -> (#531) (2 sites: scheduler.test.ts)
- issue #475 -> #475 (1 site: session-cost-usage.ts)
- bootstrap#475 -> karmaterminal/openclaw-bootstrap#475 (1 site: discoverAllSessions.test.ts; the line above already used the prefixed form, this was the laggard)
- Bug #473 fix / issue #473 -> #473 (3 sites: subagent-announce.continuation.runtime.{ts,test.ts})
- review gap on #188 -> review gap on #188 (1 site: status.test.ts; also dropped a named-prince prefix in the same test label, which separately satisfies part of #200)

Left alone (out of scope per issue):
- Pre-existing upstream openclaw/openclaw citations (openclaw#26905, openclaw#36142, openclaw#60552, openclaw#28491, openclaw#41981, openclaw#18264, openclaw#59428, openclaw#31311, openclaw#14869, openclaw#17971, openclaw#29683, openclaw#58409, openclaw#35481, openclaw#30115, openclaw#14396, openclaw#19537, openclaw#1205, openclaw#1435, openclaw#26881, openclaw#32228, openclaw#35074, openclaw#26881)  -- these were present at v2026.4.14 and resolve correctly.
- delegate-dispatch.ts:62 (`openclaw#189.`) -- already in the canonical prefixed form per the issue's exception note.

Refs #199