feat(session): add includeArchived to chat.history and sessions_history#73883
Open
CadanHu wants to merge 3 commits into
Open
feat(session): add includeArchived to chat.history and sessions_history#73883CadanHu wants to merge 3 commits into
CadanHu wants to merge 3 commits into
Conversation
Implements PR 3 of the issue openclaw#45003 archive UX roadmap (read-only path): give agents and the chat.history endpoint a way to recover context that lives in .reset.<ts> archives, without resurrecting those archives as live sessions. Changes ------- Protocol (additive, backwards-compatible): - src/gateway/protocol/schema/logs-chat.ts: ChatHistoryParamsSchema gains an optional includeArchived: boolean. Default false preserves all current behavior. Gateway helpers: - src/gateway/session-transcript-files.fs.ts: New listResetArchivesForSession(sessionId, searchDirs) lists every <sessionId>.jsonl.reset.<ts> archive sorted oldest-first. - src/gateway/session-utils.fs.ts: New readSessionMessagesIncludingArchives(sessionId, storePath, sessionFile) chains archive segments before the primary transcript with synthetic system messages tagged __openclaw.kind: "session-reset" so the Web UI / agent can render boundaries. Archive segments are parsed line-by-line; the primary segment delegates to the existing readSessionMessages so it still gets SessionManager active-branch treatment for tree-format transcripts. Gateway handler: - src/gateway/server-methods/chat.ts: chat.history reads includeArchived from params and dispatches to the new aggregator only when explicitly true. Agent tool: - src/agents/tools/sessions-history-tool.ts: SessionsHistoryToolSchema gains optional includeArchived. The tool forwards the flag verbatim to the chat.history call. Tests ----- src/gateway/session-utils.fs.test.ts (6 new tests): - Default behavior unchanged when no archives exist - Archive content surfaced when primary missing - Multiple archives chained chronologically before primary, each with a session-reset boundary marker - Empty result when neither primary nor any archive exists - Other sessions' archives are not pulled in - Empty sessionId short-circuits src/agents/tools/sessions-history-tool.test.ts (2 new tests): - includeArchived=true forwards through to chat.history - includeArchived omitted forwards as false Design notes ------------ This is the read-only path of the openclaw#45003 roadmap. It does not rename, move, or otherwise mutate archive files: archives stay archived, the user opts into visibility for one query at a time. That keeps it consistent with existing "archive lifecycle is one-way" precedent (sessions.delete has no undo) and removes the compaction-chain checkpoint and session-memory-hook double-summary risks that come with any unarchive/restore design. Refs ---- Issue: openclaw#45003 Companion PRs in the same area: - openclaw#60409 (chat.history fallback to most recent reset archive when primary is missing) — different code path, no overlap - openclaw#71537 (session-memory hook + session-logs skill archive support) — different code path, no overlap Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Contributor
Greptile SummaryAdds optional Confidence Score: 4/5Safe to merge; the one flagged issue is a non-blocking P2 about seq uniqueness that won't affect existing callers. No P0 or P1 issues found. The single P2 concern (non-unique __openclaw.seq across chained archive segments) is speculative and only matters if a consumer treats seq as a globally unique identifier, which is a new use-case introduced by this feature. src/gateway/session-utils.fs.ts — readArchivedTranscriptMessages local seq counter Prompt To Fix All With AIThis is a comment left during a code review.
Path: src/gateway/session-utils.fs.ts
Line: 213-257
Comment:
**`seq` resets to 0 for every archive segment**
`messageSeq` is declared locally inside `readArchivedTranscriptMessages` and restarts at 1 for each archive file. When `readSessionMessagesIncludingArchives` chains N archives plus the primary, the combined result contains duplicate `__openclaw.seq` values — e.g., the first message from archive-1 and the first message from archive-2 both carry `seq: 1`. If any consumer (Web UI renderer, deduplication logic, React key prop) relies on `seq` being unique across the combined list, this will produce silent bugs or duplicate-key warnings.
Consider threading a shared offset into the helper so the combined output has monotonically increasing `seq` across all segments:
```typescript
function readArchivedTranscriptMessages(filePath: string, seqOffset = 0): { messages: unknown[]; nextSeq: number } {
// ...
let messageSeq = seqOffset;
// ...
return { messages, nextSeq: messageSeq };
}
```
And in `readSessionMessagesIncludingArchives`:
```typescript
let seqOffset = 0;
for (const archive of archives) {
const { messages, nextSeq } = readArchivedTranscriptMessages(archive.path, seqOffset);
combined.push(...messages);
seqOffset = nextSeq;
// push boundary marker ...
}
```
How can I resolve this? If you propose a fix, please make it concise.Reviews (1): Last reviewed commit: "feat(session): add includeArchived to ch..." | Re-trigger Greptile |
Comment on lines
+213
to
+257
| function readArchivedTranscriptMessages(filePath: string): unknown[] { | ||
| let lines: string[]; | ||
| try { | ||
| lines = fs.readFileSync(filePath, "utf-8").split(/\r?\n/); | ||
| } catch { | ||
| return []; | ||
| } | ||
| const messages: unknown[] = []; | ||
| let messageSeq = 0; | ||
| for (const line of lines) { | ||
| if (!line.trim()) { | ||
| continue; | ||
| } | ||
| try { | ||
| const parsed = JSON.parse(line); | ||
| if (parsed?.message) { | ||
| messageSeq += 1; | ||
| messages.push( | ||
| attachOpenClawTranscriptMeta(parsed.message, { | ||
| ...(typeof parsed.id === "string" ? { id: parsed.id } : {}), | ||
| seq: messageSeq, | ||
| }), | ||
| ); | ||
| continue; | ||
| } | ||
| if (parsed?.type === "compaction") { | ||
| const ts = typeof parsed.timestamp === "string" ? Date.parse(parsed.timestamp) : Number.NaN; | ||
| const timestamp = Number.isFinite(ts) ? ts : Date.now(); | ||
| messageSeq += 1; | ||
| messages.push({ | ||
| role: "system", | ||
| content: [{ type: "text", text: "Compaction" }], | ||
| timestamp, | ||
| __openclaw: { | ||
| kind: "compaction", | ||
| id: typeof parsed.id === "string" ? parsed.id : undefined, | ||
| seq: messageSeq, | ||
| }, | ||
| }); | ||
| } | ||
| } catch { | ||
| // ignore malformed lines | ||
| } | ||
| } | ||
| return messages; |
Contributor
There was a problem hiding this comment.
seq resets to 0 for every archive segment
messageSeq is declared locally inside readArchivedTranscriptMessages and restarts at 1 for each archive file. When readSessionMessagesIncludingArchives chains N archives plus the primary, the combined result contains duplicate __openclaw.seq values — e.g., the first message from archive-1 and the first message from archive-2 both carry seq: 1. If any consumer (Web UI renderer, deduplication logic, React key prop) relies on seq being unique across the combined list, this will produce silent bugs or duplicate-key warnings.
Consider threading a shared offset into the helper so the combined output has monotonically increasing seq across all segments:
function readArchivedTranscriptMessages(filePath: string, seqOffset = 0): { messages: unknown[]; nextSeq: number } {
// ...
let messageSeq = seqOffset;
// ...
return { messages, nextSeq: messageSeq };
}And in readSessionMessagesIncludingArchives:
let seqOffset = 0;
for (const archive of archives) {
const { messages, nextSeq } = readArchivedTranscriptMessages(archive.path, seqOffset);
combined.push(...messages);
seqOffset = nextSeq;
// push boundary marker ...
}Prompt To Fix With AI
This is a comment left during a code review.
Path: src/gateway/session-utils.fs.ts
Line: 213-257
Comment:
**`seq` resets to 0 for every archive segment**
`messageSeq` is declared locally inside `readArchivedTranscriptMessages` and restarts at 1 for each archive file. When `readSessionMessagesIncludingArchives` chains N archives plus the primary, the combined result contains duplicate `__openclaw.seq` values — e.g., the first message from archive-1 and the first message from archive-2 both carry `seq: 1`. If any consumer (Web UI renderer, deduplication logic, React key prop) relies on `seq` being unique across the combined list, this will produce silent bugs or duplicate-key warnings.
Consider threading a shared offset into the helper so the combined output has monotonically increasing `seq` across all segments:
```typescript
function readArchivedTranscriptMessages(filePath: string, seqOffset = 0): { messages: unknown[]; nextSeq: number } {
// ...
let messageSeq = seqOffset;
// ...
return { messages, nextSeq: messageSeq };
}
```
And in `readSessionMessagesIncludingArchives`:
```typescript
let seqOffset = 0;
for (const archive of archives) {
const { messages, nextSeq } = readArchivedTranscriptMessages(archive.path, seqOffset);
combined.push(...messages);
seqOffset = nextSeq;
// push boundary marker ...
}
```
How can I resolve this? If you propose a fix, please make it concise.
Contributor
|
Codex review: needs real behavior proof before merge. Reviewed May 30, 2026, 12:55 AM ET / 04:55 UTC. Summary PR surface: Source +168, Tests +183, Other +8. Total +359 across 10 files. Reproducibility: unclear. The review failed before ClawSweeper could establish a reproduction path. Review metrics: none identified. Merge readiness Overall follows the weaker of proof and patch quality, so missing proof can cap an otherwise strong patch. Risk before merge
Maintainer options:
Next step before merge
Review detailsBest possible solution: Retry the Codex review after fixing the execution failure. Do we have a high-confidence way to reproduce the issue? Unclear. The review failed before ClawSweeper could establish a reproduction path. Is this the best way to solve the issue? Unclear. Retry the review first so ClawSweeper can evaluate the actual issue and fix direction. AGENTS.md: unclear because the file could not be read completely. Codex review notes: model gpt-5.5, reasoning high; reviewed against b9933b2ec119. Label changesLabel justifications:
Evidence reviewedPR surface: Source +168, Tests +183, Other +8. Total +359 across 10 files. View PR surface stats
What I checked:
Likely related people:
What the crustacean ranks mean
Shiny media proof means a screenshot, video, or linked artifact directly shows the changed behavior. Runtime, network, CSP, and security claims still need visible diagnostics. How this review workflow works
|
…udeArchived Updates the auto-generated GatewayModels.swift in both apps/macos and apps/shared/OpenClawKit to reflect the new optional includeArchived field added to ChatHistoryParamsSchema. Run: pnpm protocol:gen && pnpm protocol:gen:swift Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The previous run failed in command-execution-startup.test.ts:75 (banner-emit assertion) which passes locally 3/3 — flaky CI run. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
This was referenced Apr 30, 2026
22 tasks
masatohoshino
added a commit
to masatohoshino/openclaw
that referenced
this pull request
May 2, 2026
…reads when active is missing When the active session transcript file is missing, async read paths used by chat.history (`readSessionMessagesAsync`, `readRecentSessionMessagesAsync`) and the session title/preview lookup (`readSessionTitleFieldsFromTranscriptAsync`) returned an empty result instead of any archived content. Reset rotates the active `<id>.jsonl` to a sibling `<id>.jsonl.reset.<UTC timestamp>` and creates a new empty active file, so sessions reset before the next read became invisible through the production async API even when the archive still existed on disk. This change adds a narrow auto-fallback on the async paths only: - New `findLatestResetArchiveAsync(candidatePaths)` helper in `session-transcript-files.fs.ts` does an async `readdir` of each candidate parent directory, validates archive entries via the existing `parseSessionArchiveTimestamp(entry, "reset")` primitive, and returns only the path with the newest archive timestamp. - A small internal helper `findActiveOrLatestResetArchiveAsync` in `session-utils.fs.ts` reuses the existing active-candidate `existsSync` check first; only when no active candidate exists does it consult the new helper. - The three async functions above are re-routed through this helper. Behavior boundaries (kept narrow on purpose): - Active wins when present (no active+archive merging). - Only the single newest archive is returned (no chain aggregation, no `previousFragments`, no rolled logical-history reconstruction). - Sync paths, `findExistingTranscriptPath`, the protocol schema, the `includeArchived` opt-in, sessions_history schema, the session-memory hook, and the session-logs skill are untouched. Refs openclaw#56131 openclaw#43929 openclaw#45003 openclaw#60409 openclaw#73883
masatohoshino
added a commit
to masatohoshino/openclaw
that referenced
this pull request
May 2, 2026
masatohoshino
added a commit
to masatohoshino/openclaw
that referenced
this pull request
May 2, 2026
…hive fallback as readers Codex/clawsweeper review on PR openclaw#76134 found that `readRecentSessionMessagesWithStatsAsync` still derived `totalMessages` from `readSessionMessageCountAsync`, which kept the active-only `findExistingTranscriptPath` resolution. After the new async reader fallback, an archive-only transcript longer than the bounded tail window therefore returned the newest messages with low `seq` values and `hasMore = false`. The sessions-history HTTP first page passes `boundedSnapshot.totalMessages` into `buildSessionHistorySnapshot`, so a `nextCursor` follow-up could skip the middle archive page. This commit routes the async count and visit helpers (`readSessionMessageCountAsync` and `visitSessionMessagesAsync`) through the same `findActiveOrLatestResetArchiveAsync` resolution that the message readers use, so an archive-only transcript reports a consistent total, gives messages the right tail-end `seq` values, and exposes the archive's earlier pages via `nextCursor`. Behavior boundaries (kept narrow, same as the original PR): - Active wins when present (no active+archive merging). - Only the single newest archive is returned (no chain aggregation). - Sync paths, `findExistingTranscriptPath`, the protocol schema, the `includeArchived` opt-in, sessions_history schema, the session-memory hook, and the session-logs skill are still untouched. Tests: 4 new cases in `session-utils.fs.test.ts` cover the count and stats helpers (active-only, archive-only with archive longer than the tail window, both-exist active-priority, both-missing) and assert the returned `seq` metadata matches the archive's tail position. One new case in `sessions-history-http.test.ts` exercises the archive-only HTTP pagination path: archive a 3-message transcript, request the first `limit=2` page, follow `nextCursor`, and assert the older archive message is visible without skipping middle pages. Refs openclaw#56131 openclaw#43929 openclaw#45003 openclaw#60409 openclaw#73883
masatohoshino
added a commit
to masatohoshino/openclaw
that referenced
this pull request
May 17, 2026
…reads when active is missing When the active session transcript file is missing, async read paths used by chat.history (`readSessionMessagesAsync`, `readRecentSessionMessagesAsync`) and the session title/preview lookup (`readSessionTitleFieldsFromTranscriptAsync`) returned an empty result instead of any archived content. Reset rotates the active `<id>.jsonl` to a sibling `<id>.jsonl.reset.<UTC timestamp>` and creates a new empty active file, so sessions reset before the next read became invisible through the production async API even when the archive still existed on disk. This change adds a narrow auto-fallback on the async paths only: - New `findLatestResetArchiveAsync(candidatePaths)` helper in `session-transcript-files.fs.ts` does an async `readdir` of each candidate parent directory, validates archive entries via the existing `parseSessionArchiveTimestamp(entry, "reset")` primitive, and returns only the path with the newest archive timestamp. - A small internal helper `findActiveOrLatestResetArchiveAsync` in `session-utils.fs.ts` reuses the existing active-candidate `existsSync` check first; only when no active candidate exists does it consult the new helper. - The three async functions above are re-routed through this helper. Behavior boundaries (kept narrow on purpose): - Active wins when present (no active+archive merging). - Only the single newest archive is returned (no chain aggregation, no `previousFragments`, no rolled logical-history reconstruction). - Sync paths, `findExistingTranscriptPath`, the protocol schema, the `includeArchived` opt-in, sessions_history schema, the session-memory hook, and the session-logs skill are untouched. Refs openclaw#56131 openclaw#43929 openclaw#45003 openclaw#60409 openclaw#73883
masatohoshino
added a commit
to masatohoshino/openclaw
that referenced
this pull request
May 17, 2026
masatohoshino
added a commit
to masatohoshino/openclaw
that referenced
this pull request
May 17, 2026
…hive fallback as readers Codex/clawsweeper review on PR openclaw#76134 found that `readRecentSessionMessagesWithStatsAsync` still derived `totalMessages` from `readSessionMessageCountAsync`, which kept the active-only `findExistingTranscriptPath` resolution. After the new async reader fallback, an archive-only transcript longer than the bounded tail window therefore returned the newest messages with low `seq` values and `hasMore = false`. The sessions-history HTTP first page passes `boundedSnapshot.totalMessages` into `buildSessionHistorySnapshot`, so a `nextCursor` follow-up could skip the middle archive page. This commit routes the async count and visit helpers (`readSessionMessageCountAsync` and `visitSessionMessagesAsync`) through the same `findActiveOrLatestResetArchiveAsync` resolution that the message readers use, so an archive-only transcript reports a consistent total, gives messages the right tail-end `seq` values, and exposes the archive's earlier pages via `nextCursor`. Behavior boundaries (kept narrow, same as the original PR): - Active wins when present (no active+archive merging). - Only the single newest archive is returned (no chain aggregation). - Sync paths, `findExistingTranscriptPath`, the protocol schema, the `includeArchived` opt-in, sessions_history schema, the session-memory hook, and the session-logs skill are still untouched. Tests: 4 new cases in `session-utils.fs.test.ts` cover the count and stats helpers (active-only, archive-only with archive longer than the tail window, both-exist active-priority, both-missing) and assert the returned `seq` metadata matches the archive's tail position. One new case in `sessions-history-http.test.ts` exercises the archive-only HTTP pagination path: archive a 3-message transcript, request the first `limit=2` page, follow `nextCursor`, and assert the older archive message is visible without skipping middle pages. Refs openclaw#56131 openclaw#43929 openclaw#45003 openclaw#60409 openclaw#73883
masatohoshino
added a commit
to masatohoshino/openclaw
that referenced
this pull request
May 17, 2026
…reads when active is missing When the active session transcript file is missing, async read paths used by chat.history (`readSessionMessagesAsync`, `readRecentSessionMessagesAsync`) and the session title/preview lookup (`readSessionTitleFieldsFromTranscriptAsync`) returned an empty result instead of any archived content. Reset rotates the active `<id>.jsonl` to a sibling `<id>.jsonl.reset.<UTC timestamp>` and creates a new empty active file, so sessions reset before the next read became invisible through the production async API even when the archive still existed on disk. This change adds a narrow auto-fallback on the async paths only: - New `findLatestResetArchiveAsync(candidatePaths)` helper in `session-transcript-files.fs.ts` does an async `readdir` of each candidate parent directory, validates archive entries via the existing `parseSessionArchiveTimestamp(entry, "reset")` primitive, and returns only the path with the newest archive timestamp. - A small internal helper `findActiveOrLatestResetArchiveAsync` in `session-utils.fs.ts` reuses the existing active-candidate `existsSync` check first; only when no active candidate exists does it consult the new helper. - The three async functions above are re-routed through this helper. Behavior boundaries (kept narrow on purpose): - Active wins when present (no active+archive merging). - Only the single newest archive is returned (no chain aggregation, no `previousFragments`, no rolled logical-history reconstruction). - Sync paths, `findExistingTranscriptPath`, the protocol schema, the `includeArchived` opt-in, sessions_history schema, the session-memory hook, and the session-logs skill are untouched. Refs openclaw#56131 openclaw#43929 openclaw#45003 openclaw#60409 openclaw#73883
masatohoshino
added a commit
to masatohoshino/openclaw
that referenced
this pull request
May 17, 2026
masatohoshino
added a commit
to masatohoshino/openclaw
that referenced
this pull request
May 17, 2026
…hive fallback as readers Codex/clawsweeper review on PR openclaw#76134 found that `readRecentSessionMessagesWithStatsAsync` still derived `totalMessages` from `readSessionMessageCountAsync`, which kept the active-only `findExistingTranscriptPath` resolution. After the new async reader fallback, an archive-only transcript longer than the bounded tail window therefore returned the newest messages with low `seq` values and `hasMore = false`. The sessions-history HTTP first page passes `boundedSnapshot.totalMessages` into `buildSessionHistorySnapshot`, so a `nextCursor` follow-up could skip the middle archive page. This commit routes the async count and visit helpers (`readSessionMessageCountAsync` and `visitSessionMessagesAsync`) through the same `findActiveOrLatestResetArchiveAsync` resolution that the message readers use, so an archive-only transcript reports a consistent total, gives messages the right tail-end `seq` values, and exposes the archive's earlier pages via `nextCursor`. Behavior boundaries (kept narrow, same as the original PR): - Active wins when present (no active+archive merging). - Only the single newest archive is returned (no chain aggregation). - Sync paths, `findExistingTranscriptPath`, the protocol schema, the `includeArchived` opt-in, sessions_history schema, the session-memory hook, and the session-logs skill are still untouched. Tests: 4 new cases in `session-utils.fs.test.ts` cover the count and stats helpers (active-only, archive-only with archive longer than the tail window, both-exist active-priority, both-missing) and assert the returned `seq` metadata matches the archive's tail position. One new case in `sessions-history-http.test.ts` exercises the archive-only HTTP pagination path: archive a 3-message transcript, request the first `limit=2` page, follow `nextCursor`, and assert the older archive message is visible without skipping middle pages. Refs openclaw#56131 openclaw#43929 openclaw#45003 openclaw#60409 openclaw#73883
4 tasks
|
This pull request has been automatically marked as stale due to inactivity. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Implements the read-only path of #45003's 3-PR roadmap: give agents and
chat.historycallers a way to recover context that lives in.reset.<ts>archives, without resurrecting those archives as live sessions.This PR closes the three concrete upstream gaps identified by clawsweeper / Codex review on issue #45003 (2026-04-28T22:20:54Z):
Changes
Protocol (additive, backwards-compatible)
src/gateway/protocol/schema/logs-chat.ts—ChatHistoryParamsSchemagains an optionalincludeArchived: boolean. Defaultfalsepreserves all current behavior.Gateway helpers
src/gateway/session-transcript-files.fs.ts— newlistResetArchivesForSession(sessionId, searchDirs)lists every<sessionId>.jsonl.reset.<ts>archive sorted oldest-first.src/gateway/session-utils.fs.ts— newreadSessionMessagesIncludingArchives(sessionId, storePath, sessionFile)chains archive segments before the primary transcript with synthetic system messages tagged__openclaw.kind: "session-reset"so callers can render dividers. Archive segments are parsed line-by-line; the primary segment delegates to the existingreadSessionMessagesso it still getsSessionManageractive-branch treatment for tree-format transcripts.Gateway handler
src/gateway/server-methods/chat.ts—chat.historyreadsincludeArchivedfrom params and dispatches to the new aggregator only when explicitlytrue.Agent tool
src/agents/tools/sessions-history-tool.ts—SessionsHistoryToolSchemagains optionalincludeArchived. The tool forwards the flag verbatim to thechat.historycall.Tests
src/gateway/session-utils.fs.test.ts— 6 new tests:sessionIdshort-circuitssrc/agents/tools/sessions-history-tool.test.ts— 2 new tests verifying the param is plumbed through from tool →chat.history.Design notes
This is the read-only path of the #45003 roadmap. It does not rename, move, or otherwise mutate archive files: archives stay archived, the user opts into visibility for one query at a time. That keeps it consistent with the existing "archive lifecycle is one-way" precedent (
sessions.deletehas no undo) and avoids the compaction-chain checkpoint and session-memory-hook double-summary risks that come with any unarchive/restore design.includeArchivedis also a user-explicit opt-in (not an automatic fallback), so it is consistent with/resetsemantics: the user is asking for the recovery view, not having archived content silently re-injected into automated paths.Verification
pnpm test src/gateway/session-utils.fs.test.ts— 60/60 pass (54 originals + 6 new).pnpm tsgo:coreintroduces 0 new errors on the touched files (verified by stash-toggle baseline comparison).:18789against the local production data dir;chat.historyRPC withincludeArchived: truereturns chained archive segments withsession-resetboundary markers, while the default (false/ omitted) returns the same set as before.Related work
chat.historyfallback to most recent reset archive when primary is missing (PR 1 of the roadmap; in flight, CI green).session-memoryhook +session-logsskill archive support (touchessrc/hooks/bundled/session-memory/*, no overlap with this PR).