Commit Graph
11 Commits
Author SHA1 Message Date
Yichen Jiang 5234a40f9c refactor(llm): classify auxiliary request purpose 2026-07-22 18:59:49 +08:00
Jinhua Zhu 66e5f8eecd feat(llm): mark compaction requests with x-deepseek-harness-compact header 2026-07-22 18:35:32 +08:00
Tianyi Cui 21721bf3b1 Merge branch 'master' into fix/compaction-kv-cache-reuse
# Conflicts:
#	packages/compact/compact-basic/src/index.ts
#	packages/compact/compact-basic/src/summarizer.ts
#	packages/compact/compact-basic/tests/compact-basic.spec.ts
2026-07-21 19:45:43 +08:00
Turtle a0268d2509 fix(compact-basic): replay conversation prefix so summarization reuses KV cache
Automatic compaction fires mid-conversation, right after the loop warmed
the provider's KV cache with the last routed request. The default
summarizer then issued a separate request whose prefix shared nothing
with that warm request — a bespoke summarizer system prompt followed by
the older history flattened to one rendered transcript string — so a
differing first token invalidated the entire cached prefix and every
compaction re-processed the whole replayed history twice.

Move the compaction directive from the FRONT (a fresh system prompt) to
the END (a trailing user message), and replay the last routed request's
own system prompt, tools, message prefix, and shadowed-region messages
verbatim via session.requestHeader() + deriveEventMessage. The auxiliary
call is now a genuine prefix-extension of the warm request, so the
provider reuses the cached tokens up to the trailing instruction.

SummarizationInput carries the replayed prefix instead of a flat string;
the now-unused renderTranscript/renderContentBlocks path is removed with
its spec. Cache reuse is best-effort (head compaction guarantees a hit;
a mid-range compaction or a differently-routed summarizer forgoes it),
correctness is not.
2026-07-21 13:55:22 +08:00
Yichen Jiang b6f6c58435 Merge remote-tracking branch 'origin/master' into worktree/routed-model-compaction-policy
# Conflicts:
#	.agents/notes/implemented/architecture/2026-07-10-after-call-compaction-pressure-and-overflow-recovery.i18n.yaml
#	.agents/notes/implemented/architecture/2026-07-10-after-call-compaction-pressure-and-overflow-recovery.md
#	.agents/notes/implemented/architecture/2026-07-10-after-call-compaction-pressure-and-overflow-recovery.zh.md
#	docs/cordis-catalog/events.md
#	docs/cordis-catalog/services.md
#	docs/event-producer-consumer.md
#	examples/headless-agent/tests/harness.ts
#	examples/repl-agent/cordis.yml
#	packages/compact/compact-basic/README.md
#	packages/compact/compact-basic/src/index.ts
#	packages/compact/compact-basic/tests/compact-basic.spec.ts
#	packages/compact/compact-basic/tests/loader-composition.spec.ts
#	packages/cordis/tool-cordis/src/api-catalog.ts
#	packages/llm/README.md
#	packages/llm/llm-deepseek/src/adapter.ts
#	packages/llm/llm-pi-ai/src/adapter.ts
#	packages/llm/llm/README.md
#	packages/llm/llm/src/index.ts
#	scripts/gen-cordis-catalog.ts
#	website/zh-CN/api/harness/events.md
#	website/zh-CN/api/harness/llm.md
#	website/zh-CN/api/harness/token-meter.md
#	website/zh-CN/guide/config.md
2026-07-21 10:17:55 +08:00
Yichen Jiang cfa180c127 Resolve compaction policy per routed model 2026-07-20 15:34:00 +08:00
Tianyi Cui 3b0b0cefeb feat: implement bounded LLM request recovery 2026-07-20 03:34:19 +08:00
Tianyi Cui 7be78c7d4c Merge latest origin/master into compact-post-step-overflow-recovery
# Conflicts:
#	docs/agent-lifecycle.md
#	docs/architecture.md
#	docs/cookbook/extension-cookbook.i18n.yaml
#	docs/cordis-catalog/events.md
#	docs/cordis-catalog/services.md
#	docs/event-producer-consumer.md
#	docs/rfc/INDEX.md
#	docs/rfc/implemented/architecture/2026-07-05-reconstructable-requests.md
#	docs/rfc/implemented/architecture/2026-07-15-replay-token-meter-service.i18n.yaml
#	docs/rfc/implemented/architecture/2026-07-15-replay-token-meter-service.md
#	docs/rfc/implemented/architecture/2026-07-15-replay-token-meter-service.zh.md
#	docs/rfc/implemented/feature/2026-06-18-compaction-capability-seam.md
#	docs/rfc/implemented/feature/2026-07-07-session-prefix.md
#	packages/compact/compact-basic/README.md
#	packages/compact/compact-basic/src/config.ts
#	packages/compact/compact-basic/src/index.ts
#	packages/compact/compact-basic/src/summarizer.ts
#	packages/compact/compact-basic/tests/compact-basic.spec.ts
#	packages/compact/compact-basic/tests/compact-loop-repro.spec.ts
#	packages/compact/compact/README.md
#	packages/cordis/tool-cordis/src/api-catalog.ts
#	packages/core/agent-loop/src/loop.ts
#	packages/core/agent-loop/tests/cancel.spec.ts
#	packages/llm/llm-deepseek/src/adapter.ts
#	packages/llm/llm-pi-ai/README.md
#	packages/llm/llm-pi-ai/src/stream.ts
#	packages/llm/llm-pi-ai/tests/convert.spec.ts
#	packages/llm/llm/README.md
#	packages/llm/llm/src/index.ts
#	packages/llm/llm/tests/service.spec.ts
#	scripts/gen-doc-graphs.ts
2026-07-19 12:06:23 +08:00
Tianyi Cui f943bc8d4c Merge latest origin/master into token-meter-service 2026-07-17 22:38:46 +08:00
Hypatia May 12484104c8 feat(compact): recover context overflow (PR3 phase 2) 2026-07-15 16:50:44 +08:00
Hypatia May f038780ff6 feat(llm): add replay token metering (PR2 round 1) 2026-07-15 14:47:29 +08:00