Delete @deepseek-ai/dsh-invariants, every package's ./invariant companion
and its tests, the invariant gates and Vitest setup, the scoped-event
resolver generator, and the INVARIANT-coded listener rethrow paths.
Update docs, READMEs, Agent Notes, and add an upgrade guide.
Tool changes previously forced prompt consolidation even when the prepared
route could activate new tools through history. Rewriting the first system
message, or clearing an earlier appended prompt, reset the declaration
series and turned additions into immediate declarations.
Only routes without toolUpdate need that tool-triggered consolidation.
Capable routes append the full changed prompt and keep existing messages.
Explicit series starts, surface replacements, and unsupported prompt
routes still consolidate; retained-name definition changes still rebuild
tool declarations in Session.
Cover both update modes and reset paths, and pin the combined prompt/tool
transition through the shipped headless profile with shared schema goldens.
Log tool changes as developer messages independently of model capability,
fold committed headers and updates into Session.toolHistory(), and let the
LLM runtime construct deferred and retained declarations for routes that
declare a toolUpdate mode. DeepSeek serializes updates as system-role
tool_addition/tool_removal blocks with defer_loading and the tool-changes
beta header. An unchanged restored tool continues the series and is
re-offered through its recorded blocks instead of rewriting declarations.
Replace the obsolete rejection tests with wire assertions and add coverage
for the session fold, loop emission, route projection, config validation,
and the beta header. Chat and Trajectory changes land separately.
The 2026-07-16 explicit-turn-cancellation note still described a frozen,
detached signal reason and a two-variant cause union. The README paragraph
added here points readers at that note as the owner of the signal
lifecycle, so the stale mechanism was load-bearing.
It now records why freezing the caller's cause is not an alternative:
undici assigns `stack` from strict-mode code, so a frozen reason makes
fetch reject with `TypeError: Cannot add property stack` in place of the
abort reason.
Also names `turn/end`'s declared `TurnEndCancelCause` in the README and
takes the reviewer's helper-doc wording.
Node's fetch assigns an enumerable `stack` string onto the abort reason
object it receives, and the loop logged that same object into `turn/end`.
The undici trace became durable session data under a field no reader
declares. A transport adding a property JSON discards instead made
`Session.append` reject the ending, leaving a turn with no close.
`abortedCancelCause()` narrows the untyped `AbortSignal.reason` once and
copies the declared fields, replacing the three `as AgentCancelCause`
assertions at the two read sites and the log write.
Preserve the V3 system-prompt projection while adopting master’s per-agent frozen-message provenance and frontend performance gates. Keep benchmark workload admission and diagnostics fixes intact, and retain both migration and first-open/reopen lifecycle guarantees.
Seed the request surface watermark from the attached session so a replacement before the first resumed request consolidates retained prompt versions. An unchanged resume still continues the series. Actual-loop replacement regression fails before the fix; 53 focused tests and the agent-loop project typecheck pass.
Use the same logged per-node normalization for an empty rendering as for incapable routes and broken request series: empty each active later system node, then empty the first node. Clearing only the latest node exposes stale earlier instructions, including converter-demoted user text. Dormant empty tails neither override effective text nor produce repeated replacements.
Keep the cleared structural head and ordinary restoration semantics: a capable continuing series may append new nonempty instructions, while incapable or broken-series requests refill the head. No initial empty-head creation change, deletion event, or rollback machinery.
Tests: capable/plain clearing after three active prompt versions fails before the fix and passes afterward; repeated clear and seeded resume remain empty without duplicate system events, restoration includes only new instructions, pi-ai conversion has no stale prompt, and source-event replay reconstructs every request. Final 395 loop tests pass with exact agent.ts/runtime-context.ts coverage at 100% all metrics; source tsc and focused lint pass. Fix new test discriminant narrowing discovered by focused compiler audit. Update EN/ZH consumers and owning rationale, seven pairs verified; parent owns broad gates and generated catalog/SDK/GUI updates.
Every same-step attempt resolves its bound route and reconciles the accepted rendered prompt against the current surface before deriving history. Preserve assembly, pre-step admission and entered users exactly once; compaction retries must not resurrect an older surviving prompt.
Normalize a broken request series to the current head plus dormant empty later system nodes, even if later versions survive or the effective text is unchanged. This removes the later-survivor exception without a delete operation and avoids appending instructions after already-admitted users on retry. Capture surface generation after reconciliation so subsequent ordinary retries do not emit phantom series headers.
Tests: both compaction cases fail before the fix (stale head, or stale later survivor), then pass with real loop/MockAdapter, pi-ai conversion and source-event reconstruction. Assert two retries keep one assembly/pre-step/user admission, preserve rendered text across section changes, and log one series boundary. Add unchanged-text explicit/tool series regressions. 391 loop tests with exact agent.ts/runtime-context.ts coverage pass at 100%; final admission spec has 11 passing cases. Source tsc and focused lint pass. Update EN/ZH implementation and owning rationale plus lifecycle retry documentation.
Resolve agent/request and prepareCall inside the accepted open step before committing system and user input. Preserve prompt-assembly model selection, bind capability and dispatch to the same prepared adapter, then log the envelope and derive the request synchronously. Preparation cancellation leaves a balanced empty step with no admitted input.
Normalize incapable-route retained prompt versions with logged per-node empty replacements and a current head, including unchanged rendering and resumed history. Dormant empty tails do not override the effective prompt. No new event or delete operation.
Tests: 389 agent-loop cases pass with exact agent.ts/runtime-context.ts coverage at 100% for all metrics; tsc -b packages/core/agent-loop/tsconfig.json passes. Seven actual-loop MockAdapter cases exercise pi-ai conversion, source-event reconstruction, both route transitions, resume, cancellation barriers, and bound selection. Stale-capability negative control fails all three transition regressions. Update EN/ZH request visibility, lifecycle generator, and owning rationale. Broad catalogs, recorded SDK/GUI and aggregate gates remain parent-owned.
Consolidate the existing in-history capability, loop, UI and artifacts into one baseline. Preserve the reviewed tree so each independently verified correction has a subsequent rationale-rich commit.
Cause: SystemPromptProjection skipped the first empty rendered prompt. The initial admitted user then occupied surface node zero, so a later nonempty prompt appended behind user history. Routes without in-history system support lost the leading system role; pi-ai demotes a non-leading system message to user content.
Fix: append the initial system node even when its content is empty. The existing loop commit order reserves node zero before admitted user messages; later prompt text replaces that node. Empty content still derives to no wire message. Keep retained-node replacement, clearing, multi-system handling, and pi-ai conversion unchanged; this addresses only the reviewed PR3476 initial-empty finding, not PR3483.
Tests: added initial-empty projection and two-turn loop regressions for empty wire output, reserved surface head, later leading system role, replacement intent, and series header. Negative control failed before the source fix. Focused projection/runtime-context/loop/request-reconstruction/session-surface/pi-ai-context suites passed 176 tests; exact runtime-context.ts coverage is 100% statements, branches, functions, and lines. test:docs passed all 15 gates. Updated README EN/ZH, architecture map and owning architecture note; recorded all three translation pairs. Broad doc-sync/lint stopped at parent request for combined-layer validation. No normalize.ts conflict-comment edit.
Consolidate the representation-change PR and its rebase reconciliations into one baseline. Preserve the exact tree and keep the in-history feature in the dependent PR. Follow-up fixes remain separate.
The persistence seam is now create/open/stat/list returning per-session
SessionHandles (read/append/flush/close); every log read and write flows
through the owning handle. The seam package exports only the service and
handle contracts, consumer-visible errors, and pure durable-data
validation helpers; each backend owns its complete storage runtime, and
the shared contract suites pin equivalent observable behavior. The
backend routes published sessions' live events by id into the active
write handle; agent-loop only acquires, seeds, and closes the handle.
Resume appends interruptedTurnClosers through its write handle;
session-query owns the revision-keyed cold cache. Legacy-only surfaces
are removed in the same swap: locate/readRaw/supportsRawArtifacts, the
legacy event-shape read migration, zstd torn-frame salvage,
DSH_SESSION_JSONL, and hook transcript_path population; a torn final
zstd frame is discarded whole; the session-list cold blank probe returns
on stat metadata (eventCount derived from the last physical row,
sizeBytes). The WebUI ZIP export serializes the logical log from a read
handle, so both backends export identically.
Refs #3245
Tianyi Cui