Extracts the 25 repeated NewCompositeRetrieveEngine call sites across
seven services into two factory functions with tenant-ownership
verification: CreateRetrieveEngineForKB (synchronous) and
CreateRetrieveEngineFromPayload (async task handlers). Promotes
GetByStoreID from concrete-only to interfaces.RetrieveEngineRegistry.
Extends KBDeletePayload / IndexDeletePayload with an omitempty
VectorStoreID snapshot.
Design highlights
- Factory verifies tenant ownership of the resolved store
(defense-in-depth) so a gap in PR3 validation or a tampered Asynq
payload cannot cross tenants. Cross-tenant attempts return
ErrVectorStoreForbidden with a structured log entry.
- Sentinel errors (ErrTenantInfoMissing, ErrVectorStoreNotFound,
ErrVectorStoreForbidden) let async handlers classify non-retryable
failures as asynq.SkipRetry.
- nil and empty-string pointers for vectorStoreID normalize to
"unbound" so callers never send an empty UUID to GetByStoreID.
- resolveBoundEngine constructs engineInfos directly (with
slices.Clone of Support()) instead of going through
NewCompositeRetrieveEngine. The latter only reads from the
byEngineType map and cannot reach DB stores in byStoreID -- multiple
stores can share the same engine type. Unbound fallbacks still reuse
NewCompositeRetrieveEngine so the tenant effective-engines path is
byte-for-byte unchanged.
- DeleteKnowledgeBase now reads the KB before soft-delete so the
enqueued KBDeletePayload carries a VectorStoreID snapshot; once the
row is soft-deleted GORM's default scope hides it. Cost: +1 SELECT
per KB-delete request (a rare admin operation).
- enqueueIndexDeleteTask gains a vectorStoreID *string parameter that
the single caller (tag.go) populates from the owning KB. The worker
(ProcessIndexDelete) validates ownership via the factory.
- cleanupKnowledgeResources loads the KB so bound-store cleanup routes
correctly. If the load fails it warns and falls back to tenant
effective engines; orphan vectors become observable in logs instead
of disappearing silently.
- DeleteKnowledges (batch) stays on the tenant effective-engines path
-- batches can span stores and multi-store fan-out is PR4 scope.
Noted inline.
Merge safety
All knowledge bases have VectorStoreID = NULL at merge time (introduced
nullable in PR1, Tencent/WeKnora#994). Every factory call falls back to
NewCompositeRetrieveEngine(registry, tenantInfo.GetEffectiveEngines()),
which is the existing code path for these KBs. Async payloads from
before this change decode to VectorStoreID = nil (omitempty) and take
the same fallback.
Testing
- go build ./...
- go vet ./...
- go test -race -count=1 ./internal/application/service/retriever/...
covers unbound (nil/empty), bound, cross-tenant, unregistered store,
ownership infra error, legacy payload shapes (missing field/null/
empty string/nil), tampered payload, and parallel invocation.
- grep -r NewCompositeRetrieveEngine internal/application/service
returns zero results outside factory.go/composite.go.
Refs: Part of Tencent/WeKnora#993. Depends on Tencent/WeKnora#994.
JoinByOrganizationID no longer delegates to JoinByInviteCode for direct
join, matching the product copy that searchable spaces need no invite
code. Extract joinAsViewerWithChecks for shared member-limit logic.
Co-authored-by: Cursor <cursoragent@cursor.com>
MiMo and DeepSeek V3.2/V4 reject multi-turn requests in thinking mode
when the prior assistant message lacks reasoning_content with HTTP 400:
"The reasoning_content in the thinking mode must be passed back to the API."
Agent's ReAct loop is the worst-case scenario — every round produces tool
calls, exactly the case DeepSeek's docs specify reasoning_content MUST
participate in subsequent context.
Plumb reasoning_content through the full assistant-message round-trip:
- chat.Message / types.ChatResponse / types.AgentStep gain a
reasoning_content field (AgentStep persists via the existing
Message.AgentSteps jsonb column, no migration needed).
- streamLLMToEventBus accumulates reasoning chunks into
result.ReasoningContent and surfaces it on the round's ChatResponse.
- engine.runReActIteration writes it onto AgentStep so cross-turn replay
preserves it.
- observe.appendToolResults attaches it to the same-turn assistant
message; agent_history.buildAssistantHistoryMessages does the same on
cross-turn replay.
- RemoteAPIChat.ConvertMessages forwards it on assistant turns to
openai.ChatCompletionMessage.ReasoningContent (already supported by
go-openai); providers that don't recognize the field ignore it.
Tests cover the three boundaries: ConvertMessages serializes it for
upstream, appendToolResults preserves it within the same turn, and
buildAssistantHistoryMessages replays it across turns.
Scope is intentionally limited to Agent mode — KnowledgeQA's chat
pipeline and Anthropic's signed thinking_blocks are separate fixes that
require schema changes (rendered_content / thinking_blocks columns).
Fixes#1302
Previously the Langfuse timeline for a knowledge-chat request only
showed generations (chat.completion, embedding.embed, rerank), so the
work happening between them — query setup, vector/keyword search,
result merging, prompt assembly — appeared as unexplained gaps.
Add spans around the pieces that fill those gaps:
* qa.setup wraps request-time KB / model / search-target resolution
before the pipeline event loop starts, accounting for the visible
delay before the first chat.completion.
* pipeline.<event> wraps each pipeline stage's eventManager.Trigger
call so generations inside a stage nest under it. CHAT_COMPLETION_
STREAM is intentionally skipped because its OnEvent returns as soon
as the streaming goroutine starts; a stage span would always finish
before the chat.completion.stream generation it nominally parents.
* retrieve wraps the actual vector + keyword retrieve call inside
HybridSearch, exposing the DB round-trip that previously sat invisibly
between embedding generations and rerank.
* web_search wraps the external web-search HTTP call when enabled.
Quick-answer (RAG) agents can now configure a separate chat model for
the query-understanding stage (rewrite + intent classification),
decoupling it from the main conversation model so users can route the
lightweight rewrite call to a cheaper / faster model.
- Add CustomAgentConfig.QueryUnderstandModelID and plumb it through
PipelineRequest and ChatManage.Clone.
- query_understand plugin prefers QueryUnderstandModelID on the
text-only path; falls back to ChatModelID if the configured model
cannot be resolved (with a warn log). Multimodal path is unchanged
to keep vision-capable model selection intact.
- AgentEditorModal exposes a ModelSelector under the existing Query
Rewrite block; empty means reuse the main chat model.
- Add i18n strings (queryUnderstandModel / placeholder / desc) for
zh-CN, en-US, ko-KR, ru-RU.
- Extend Go SDK AgentConfig with the new field.
The legacy in-pipeline DuckDB SQL data-analysis stage used to run on every
quick-answer RAG request whose retrieved chunks included a CSV/Excel file,
adding one extra LLM round-trip (~3s) to generate a SQL query that most
plain Q&A users never wanted. There was no way to disable it.
Introduce a per-agent DataAnalysisEnabled flag (default off), wire it
through PipelineRequest, and gate the DATA_ANALYSIS stage on it. Surface
the toggle in the agent editor for quick-answer agents with at least one
knowledge base attached.
Refs: https://github.com/Tencent/WeKnora/issues/1244
Add three new optional filters to the document list under a knowledge base
detail page — parse status, source/channel, and updated time range — and
rework multi-select to no longer cause the card title to jitter on hover.
Backend
- Introduce types.KnowledgeListFilter to aggregate optional filter dimensions
(tag, keyword, file_type, parse_status, source, updated_from/to) and switch
ListPagedKnowledgeByKnowledgeBaseID (repository/service/interface) to accept
it instead of a growing positional parameter list.
- The ListKnowledge HTTP handler accepts new parse_status, source, start_time
and end_time query params; time params accept RFC3339, "YYYY-MM-DD HH:MM:SS"
and "YYYY-MM-DD". The repository routes source="manual"/"url" onto the type
column to stay consistent with file_type semantics; other source values match
the channel column.
- Update the four other callers (agent_service, initialization) to pass an
empty filter struct, preserving prior behavior.
Frontend
- Add three controls in the doc-filter-bar (status select, source select,
date-range picker with future-date disabled) wired through getKnowled /
listKnowledgeFiles into the new backend params.
- Replace the hover-triggered card checkbox with an explicit "批量管理" mode
(mirrors the session list UX): in card view the checkbox only renders while
batch mode is on, entered via the per-card "..." menu; the list view keeps
its leading checkbox column. Switching from list to grid auto-enables batch
mode when something is already selected, so the selection stays visible.
- DocumentBatchBar now stays open whenever batch mode or selection > 0, and
its "取消选择" button both clears the selection and exits batch mode.
API surface sync
- Regenerate Swagger artifacts (docs/docs.go / swagger.json / swagger.yaml).
- Update docs/api/knowledge.md with the new query parameters.
- Add backward-compatible ListKnowledgeWithFilter + KnowledgeListFilter to the
Go SDK; the existing ListKnowledge keeps its signature.
i18n
- New filter labels in zh-CN / en-US / ko-KR / ru-RU; reuse existing
menu.batchManage / batchManage.cancel for the multi-select strings.
When image bytes for a multimodal task cannot be read via FileService
(e.g. tenant.StorageEngineConfig.MinIO is empty while the image was
saved using the global MINIO_* env vars), the previous code fell back
to the HTTP downloader, which rejected the provider:// URL with
"unsupported URL scheme". The asynq handler then returned an error,
asynq retried until exhaustion, and the per-knowledge "pending images"
counter was never decremented — leaving the document stuck in
"processing" indefinitely (issue #1282).
Changes:
- ImageMultimodalService now holds a default FileService and
resolveFileServiceForPayload falls back to it when the tenant-scoped
storage config cannot produce a usable service, mirroring the
write-side fallback in knowledgeService.resolveFileService.
- Extract readImageBytes: provider:// URLs are read exclusively via
FileService and never handed to the HTTP downloader.
- On unrecoverable read failure for a single image, log and skip that
image but still call checkAndFinalizeAllImages so the parent
knowledge can progress to post-processing.
Fixes#1282
Previously the custom agent editor hid the rerank model field when no
RAG-type knowledge base existed in the configured scope, but the
not-ready check and runtime hard-failed whenever knowledge_search was
in allowed_tools. Users with wiki-only or empty scopes saw "Rerank
Model required" warnings they could not resolve, and "All knowledge
bases" agents broke later if a RAG-type KB was added.
- Backend: when knowledge_search is enabled, fall back to the tenant
default rerank model (ConversationConfig.RerankModelID) before
erroring out, matching the built-in agent behaviour.
- Editor: always show the rerank field once a KB scope is selected;
only mark it required (red *) when the scope contains a RAG KB, with
a hint explaining the tenant-default fallback.
- Editor: only render rerank top_k / threshold sliders when a rerank
model is actually selected.
- Input field: drop the eager "missing rerank" not-ready reason; the
backend is now the single source of truth for rerank availability.
- i18n: add agent.editor.rerankModelOptionalHint across all locales.
Quick-answer agent mode retrieves purely through vector/keyword chunk
search and ships with no `allowed_tools`, so the existing capability
filter (which only reads from `allowed_tools`) let wiki-only KBs through
in every entry point. End result: users could @-mention, select, and
receive suggested questions from wiki-only KBs in quick-answer mode,
but the underlying retrieval always returned empty.
Treat "RAG-only" as an implicit property of `agent_mode = quick-answer`
and union it with the tool-derived filter. The same predicate is now
used everywhere the user can pick or be steered toward a KB:
Backend
- `tools.DeriveKBFilterForAgent` / `KBSatisfiesAgentRequirements`
layer the implicit quick-answer requirement on top of tool derivation.
- `ListKnowledgeBases`, `SearchKnowledge` (shared-agent `@file`),
`resolveKnowledgeBasesFromAgent` (chat runtime), `/search` IM command,
and `GetSuggestedQuestions` now all use the agent-mode-aware variant.
- `GetSuggestedQuestions` also skips the wiki-page fallback for
quick-answer agents to cover the `selected` / explicit-kb-ids paths
where a wiki-only KB could still slip through.
Frontend
- `deriveKbFilterForAgent` / `kbSatisfiesAgentRequirements` mirror the
Go helpers.
- `@` mention dropdown (`Input-field.vue`) uses the new helper.
- Agent editor's "specified KB" picker (`AgentEditorModal.vue`) grays
out wiki-only KBs for quick-answer agents with a tooltip, and the
pre-save warning fires for quick-answer mode too.
- i18n: add `agentEditor.agentType.kbMismatch.quickAnswer` across all
four locales.
Agent mode does not persist `rendered_content` for user messages, so
when the next turn's history was rebuilt from DB, attachments uploaded
in prior turns disappeared — the model only saw the raw query plus the
prior assistant reply, breaking follow-up questions that referenced
the file (e.g. "what is in there?").
Reconstruct the attachment prompt from the stored `Attachments` column
when `RenderedContent` is empty, mirroring how image captions are
already replayed. KnowledgeQA turns (which do persist
`RenderedContent`) are unaffected and won't get attachments injected
twice.
Refs #1237
The Agent ran a parallel context cache (Redis/in-memory) on top of the
messages table to feed multi-turn history into the LLM. That dual-write
caused subtle drift (e.g. compression diverging from DB, system-prompt
swaps lost on restart) and required a separate ClearContext path on
session/IM clear.
Make the messages table the single source of truth:
- Add service.LoadAgentHistory: rebuilds chat.Message history per turn
from the persisted messages, expanding AgentSteps into proper OpenAI
assistant_with_tool_calls + tool messages and replaying the canonical
final answer (with <think> blocks stripped). final_answer tool calls
are filtered to avoid duplicating the trailing answer.
- Make AgentEngine stateless across turns: drop ContextManager / sessionID
cache plumbing from the engine, agent service, and CreateAgentEngine
signature. The engine only uses sessionID for logging/event emission.
- Wire AgentQA to load history from DB on demand using HistoryTurns
(default 5) when MultiTurnEnabled, otherwise run with empty history.
- Delete the llmcontext package (ContextManager interface, Redis/memory
storage, factory) and the SessionService.ClearContext API path; IM
/clear and session message clear no longer need to invalidate cache.
Behavior preserved: KnowledgeQA-mode replay is unchanged (turns with
empty AgentSteps just produce the single canonical assistant message),
and Agent-mode turns now consistently see prior tool calls and results.
Add an opt-in human approval gate so Agent runs pause before executing
MCP tools that operators flag as dangerous, surface an approval card in
the chat UI, and only resume after the user approves (optionally with
edited args) or rejects.
Backend
- New mcp_tool_approvals table + repo/service to mark per-tool approval
required (PG migration 000042 + sqlite init).
- approval.Gate coordinates RequestAndWait / Resolve with sync.Once
delivery, configurable timeout, and Redis Pub/Sub fan-out so multi-
replica deployments work without sticky sessions.
- MCPTool.Execute integrates the gate; uses a round-level ApprovalCtx
(without the per-tool 60s timeout) for the wait, and re-derives a
fresh 60s exec ctx after approval so CallTool keeps a full window.
- New SSE response types (tool_approval_required / _resolved) and
EventBus events plumb approval state to AgentStreamDisplay.
- REST: list/set per-tool approval flag, resolve pending approval.
- Configurable via agent.tool_approval_timeout_seconds (yaml) or
WEKNORA_AGENT_TOOL_APPROVAL_TIMEOUT env (accepts seconds or Go
duration).
Frontend
- MCP settings: per-tool "require approval" switch on the test panel.
- Chat: ToolApprovalCard renders the pause point with editable JSON
args, validation feedback, mm:ss countdown that turns warning/danger
near deadline, and a resolved state that retains context.
- i18n strings added for zh-CN / en-US / ko-KR / ru-RU.
Docs
- docs/zh/mcp-approval.md covering behavior, config, API, deployment
considerations (Redis cross-instance, restart limitations).
- repo: drop r.db.Debug() from FindSimilarPages — it was dumping every
trigram probe's SQL+args (per-alias, per-item) into production logs.
- wiki_ingest dedup: fix Printf format string ("selected for %d new
items" had two args), and harden validMerge against un-prefixed
slugs whose strings.Index returned -1 and silently passed the type
check.
- wiki_ingest_batch: drop the duplicated loggedBatchSize/MapPar/
ReducePar assignments.
- asynqdl: record the real attempt count (retried + 1) on the dead
letter row instead of a hard-coded 0; tighten payloadProbe to the
set of field names with consistent semantics across payloads
(drop source_id/target_id/target_kb_id which differ by task type).
- asynqdl tests: update for the trimmed probe and assert FailCount=0
outside an asynq worker ctx so the semantics stay pinned.
The wiki ingest post-process pipeline OOM'd and ran for hours on KBs
with ~40k documents. The dominant tail was a per-batch ListAllPages
that pulled every page (multi-MB content blobs) into Go memory, plus a
24h-TTL Redis pending list whose data could be evicted before a long
serialized backlog drained. dedup did O(P × N) Jaccard scoring in Go
on every batch. Lint loaded the full graph. The index page kept the
entire wiki directory in its content column and rewrote the TOAST
chunk on every ingest. None of these survive at 4w docs.
This change reworks the write path end to end and pulls the durable-
queue + dead-letter primitives out of wiki and into shared
infrastructure that every asynq task type now benefits from.
- migration 000041: task_pending_ops + task_dead_letters tables, plus
three GIN indexes on wiki_pages (source_refs jsonb_path_ops,
source_refs text fallback, lower(title) trgm).
- TaskPendingOpsRepository + TaskDeadLetterRepository in
internal/application/repository/task_queue.go: cursor-paginated
list, atomic IncrFailCount via UPDATE…RETURNING, dedup-key scoped
delete that refuses empty keys so a buggy caller can't wipe a KB's
queue.
- internal/middleware/asynqdl: writes a task_dead_letters row when an
asynq task exhausts its retry budget. Payload-agnostic — a small
probe struct extracts TenantID + scope hints across every existing
payload type, so summary:generation / image:multimodal /
faq_import / kb:clone / etc. all dead-letter without per-handler
code. Best-effort: an insert failure never masks the underlying
task error. Installed in router/task.go before the langfuse mw so
it sees raw errors.
- Pending queue moved Redis → PG (no TTL, restart-durable). Redis
keeps just the active-batch lock and the delete tombstone.
- In-batch retry budget (wikiMaxFailRetries=5) tracked via
pendingRepo.IncrFailCount; over the cap the row is moved to
task_dead_letters with task_type=wiki:ingest, related_id=knowledge
id. Asynq retries (10) are handled by the new middleware.
- Removed the per-batch ListAllPages. WikiBatchContext now carries
lazy fetcher closures (SlugTitleMany, SummaryByKnowledgeID) with
mutex-protected caches; reduce reaches for titles / summaries on
demand instead of pre-loading the whole KB.
- getExistingPageSlugsForKnowledge now uses ListSlugsBySourceRef,
which the new GIN index on source_refs serves as a Bitmap Index
Scan instead of a sequential text LIKE.
- Dedup pre-filter uses pg_trgm via FindSimilarPages
(idx_wiki_pages_title_trgm). For each new entity/concept (and each
of its aliases) we ask the DB for the top-K trigram-similar
existing pages and union the results — bounded prompt size, no Go-
side O(P × M) loop. Small KBs (≤25 entities) bypass the pre-filter.
- cleanDeadLinks / injectCrossLinks are scoped to the batch's
affected slugs (a few dozen) instead of every page in the KB. Both
use the new lite ListBySlugs / ExistsSlugs repo methods so they
pull only slug + title + outlinks, not full content.
- Slug fuzzy resolve (slug_fuzzy.go): when the LLM emits
[[bad-slug|display]] the cleanup path now tries display-text
reverse lookup → hyphen/case normalized equality → char-bigram
Jaccard ≥ 0.8 before stripping. Recovers the common pinyin-word-
break drift case ("shang-hai-tower" vs "shanghai-tower") in place
instead of replacing the link with plain text.
- rebuildIndexPage uses ListByTypeRecent(200) for the first-time
intro and drops the full DocumentSummaries blob from the
incremental update prompt, so its context stays bounded regardless
of KB size.
- Concurrency tunables surfaced in WikiConfig: IngestBatchSize /
IngestMapParallel / IngestReduceParallel, with sensible defaults
via OrDefault helpers. scheduleFollowUp drops to ProcessIn(0) so
follow-ups don't waste asynq retry slots bouncing on the active
lock.
- RunLint walks pages via the new ListPagesCursor in 200-page
windows and computes the live-slug set with a one-column
ListAllSlugs Pluck instead of a Limit:0 GetGraph that materialized
every node + edge. Memory is now bounded; 40k pages walks in
constant ~4MB.
- 14 GORM tests for both repos against an in-memory SQLite mirror of
the production DDL (task_queue_test.go).
- 6 tests for the dead-letter middleware covering retry budget
detection, payload-agnostic scope inference, error truncation, and
repo-failure isolation.
- 7 tests for the slug fuzzy resolve helper covering the three
resolution stages, display-text priority over normalized equality,
bigram fallback acceptance, and rejection of unrelated slugs.
- Existing wiki_ingest / wiki_lint / wiki_page tests updated to the
new fetcher / cursor APIs.
When a doc's entity/concept page generation hit a transient LLM error in
the reduce phase, the page never got written, but the doc's summary page
(generated in parallel with the slug list baked into its content) was
already persisted with [[entity/foo|name]] links pointing at the missing
page. The wiki log feed also surfaced the failed slugs as clickable
entries that 404'd. cleanDeadLinks only ran for retract batches, so the
debris persisted indefinitely.
reduceSlugUpdates now reports addition-path failures back to the batch
driver, which collects them and:
- rewrites this batch's summary pages, replacing dead [[slug|display]]
refs with plain display text (or a humanized slug tail);
- filters dead slugs out of docResults[].Pages before the wiki log
entries are flushed;
- broadens the cleanDeadLinks trigger to fire whenever any page was
touched or any slug failed, as a long-tail safety net.
Pure text replacement; no extra LLM calls. Failed slugs are picked up
naturally on the next ingest of the same document via the existing slug-
continuity rules, so the system self-heals without manual reingest.
Avoid treating Redis pending-length read failures as empty queues, reset stale fail-count keys on fresh ingest/drop paths, and reduce wiki ingest MaxRetry to a moderate shared constant to limit queue churn.
Fixes O(n²) write amplification during wiki ingest on large KBs. Previously
every ingest/retract op re-wrote the single `slug='log'` row end-to-end and
every batch re-wrote the entire `slug='index'` directory markdown. On a 40k-
doc KB the log row grew to tens of MB and the index row to several MB, so
each batch triggered giant TOAST updates that dominated ingest wall time.
Log: new `wiki_log_entries` event table (`id DESC` indexed per KB) replaces
the single TEXT row. Batch ingest now collects entries and flushes them
once per batch via `AppendBatch`. Each entry stores `pages_affected` as
JSONB `[{slug,title}]` so the UI can render real titles; custom Scan falls
back to legacy `[]string` so older rows still deserialize.
Index: `wiki_pages[slug=index].content` keeps only the LLM-generated intro
(a few KB). The directory is now served by a structured paginated API
(`GetIndexView`) that reads `slug/title/summary` per type with cursor
pagination, so the agent and the frontend only pull the slice they need.
`RebuildIndexPage` degrades to a no-op; agent `wiki_read_page('index')`
synthesizes a small top-K overview and points callers at `wiki_search`.
Ingest resilience: LLM calls wrap with 3-attempt exponential backoff on
transient errors (5xx/408/429, transport resets/timeouts). Summary/extract
failures now bubble up so the batch's failed-op requeue path runs instead
of silently dropping the doc.
Frontend: sidebar Index/Log entries switch to dedicated views. Index view
streams intro → Summary → Entity → Concept → Synthesis → Comparison via
IntersectionObserver (with a nextTick re-check so small KBs still load
every section). Log view uses cursor pagination; pages_affected renders
titles with slug tooltip.
- new migration: migrations/versioned/000040_wiki_log_entries.{up,down}.sql
- tests: log repo pagination + legacy Scan, ListByTypeLight windowing,
renderIndexOverviewForAgent output, isTransientLLMError classifier
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Root cause: requeueFailedOps unconditionally RPush'd every failed op back to
wiki:pending, causing unbounded growth when a document consistently triggers
LLM timeouts. Observed: 553 entries in wiki:pending:08134644 (one KB), mostly
duplicates for ~5 unique documents that each timed out on every batch cycle.
Fix:
- Add wikiFailCountKeyPrefix ("wiki:failcount:") and wikiMaxFailRetries (5)
constants in wiki_ingest.go
- requeueFailedOps now atomically Incr(wiki:failcount:{kbID}:{knowledgeID})
before each RPush; skips re-queue (drops op permanently) if count > 5
- Expire the fail-count key at wikiPendingTTL (24h) on every update
- wiki_ingest_batch.go: Del the fail-count key at ingestSucceeded++ so
transient errors don't permanently burn through a document's retry budget
Evidence of problem severity: 5 persistently-failing docs × ~110 requeue
cycles each = 553 pending entries; had to use Lua atomic dedup to recover.
Behaviour after this fix:
- First 5 failures → normal retry (re-queued, follow-up batch scheduled)
- 6th failure → logged as WARN "dropping op … after 6 failures (limit 5)",
not re-queued; prevents the feedback loop
- If document is fixed upstream and re-ingested via EnqueueWikiIngest, the
op is a fresh entry with no existing fail-count key → full 5-retry budget
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Root cause: when multiple wiki:ingest tasks queue up for the same KB,
they all hit ErrWikiIngestConcurrent and retry every 15 s. With
MaxRetry=10 (150 s window), any batch taking longer than ~2.5 min
caused pending tasks to exhaust retries and land in archived.
Confirmed via Redis inspection: archived wiki:ingest task had
retry=10, retried=10, error="concurrent wiki task active".
Two-part fix:
1. Early-exit at lock conflict when pending list is empty.
If another batch is active and the pending list is already empty,
there is nothing left to process — the concurrent task will handle
everything. Return nil (success) immediately instead of burning
through retry slots on a guaranteed no-op. Only retry when the
pending list still has items, i.e., we have real work to do.
2. Increase MaxRetry 10 → 25 across all four enqueue sites.
25 × 15 s = ~6 min window; accommodates large KB batches and the
60 s orphan-lock expiry with generous headroom. Combined with the
early-exit above, redundant tasks now consume 0 retries instead of
10, so the remaining budget is entirely available for genuine
concurrency conflicts.
No API or schema changes.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Attachments uploaded from the UI were correctly parsed by the handler and
set on QARequest.Attachments, but the Agent path dropped them before
calling engine.Execute. Only the KnowledgeQA pipeline injected them via
Attachments.BuildPrompt(), so in "smart reasoning" (Agent) mode the model
never saw the uploaded file and answered as if no attachment existed.
Append Attachments.BuildPrompt() to agentQuery alongside QuotedContext,
mirroring chat_pipeline/into_chat_message.go. This keeps the engine
signature unchanged and matches the existing KnowledgeQA behavior.
Wikis with tens of thousands of pages used to crash the browser trying
to render the entire link graph at once (30MB+ JSON, 100k+ SVG elements).
This change moves the graph viewer from "fetch everything, render
everything" to "fetch a slice, expand on demand" with several matching
UX improvements.
Backend
* GET /wiki/graph now accepts `mode` (overview | ego), `center`, `depth`,
`types`, `limit` query params. Default is overview top-500 by
link_count, capped at 2000. Response includes a Meta object so the
frontend can render a truncation hint and drive the UI.
* Pure helper `computeGraphSubset` extracted for testability; six unit
tests cover overview truncation / type filter / ego BFS / missing
center error.
* WikiLintService passes Limit=0 (uncapped) so link integrity checks
still walk every page.
* Repo `Search` adds CASE-based relevance ranking (title 4, slug 3,
summary 2, content 1) so full-text results put the most obvious
matches first instead of whatever was updated most recently.
Frontend
* Graph viewer fetches overview on entry; ego pivot on double-click,
search, wiki-link click, URL ?slug=, global issues jump.
* Shift+click or hover ⊕ button blooms neighbors onto the current
canvas additively. Bloom tracks generations and LRU-evicts oldest
when total exceeds 1500 nodes; ego center / selected node / latest
anchor are always protected.
* "Grow frontier (N)" legend action expands every dashed-ring node in
one click (6-way concurrency) while skipping Index/Log super-nodes
that would otherwise dump the whole wiki onto the canvas.
* Node dashed expansion ring + drawer "X/Y neighbors shown" hint tell
users which nodes still have neighbors to load.
* Search dropdown uses remote full-text search (debounced, sequence-
numbered to drop stale responses) and falls back to the overview
top-500 snapshot when the input is empty.
* Type filters now round-trip to the server so top-N is always
computed from the user's active type set rather than hiding nodes
client-side and shrinking the view.
* Status card replaces the cramped "centered on X · N hops · M nodes"
line with a structured focus/overview summary, resolving slugs to
page titles.
* Help popover in the legend lists every canvas shortcut.
i18n entries added for zh-CN / en-US / ko-KR / ru-RU.
Closes#1113
The graph extraction pipeline relies on a strict markdown fence regex to
pull JSON out of LLM responses. In production the regex misses ~84% of
real-world responses, in three ways:
1. The LLM hits max_tokens mid-output and produces no closing fence.
2. The opening fence is malformed or surrounded by prose, so the
non-greedy regex fails to anchor a match.
3. The model returns raw JSON with stray backticks but no real fence.
In every case extractContent fell through to returning the raw text,
which then failed json.Unmarshal with errors like
"invalid character '` + "`" + "' looking for beginning of value".
This change keeps the existing happy path untouched and adds a
conservative recovery step in the default branch of extractContent:
- If an opening ``` is present, take everything after it, drop a
likely language tag on the first line, cut at any trailing closing
fence, and trim stray backticks/whitespace.
- Otherwise, look for an outermost JSON object/array in the text using
a small bracket-balanced scanner that respects string literals, so
embedded {} or [] inside JSON strings don't confuse it.
- Only fall back to the original raw-text behavior when neither
strategy yields anything plausible.
The recovery helpers (stripFencesAndExtract, extractJSONLike,
isLikelyLanguageTag) are package-private and have table-driven tests in
extract_entity_test.go covering the three failure patterns described in
the issue, plus the previously-working fenced and bare-JSON shapes to
guard against regressions.
internal/application/service/knowledge.go had grown to 9883 lines /
149 functions spanning CRUD, document processing, summary/question
generation, clone/move, FAQ (CRUD+import+index+export), wiki cleanup,
multimodal image, and file utilities. Navigating it, or diffing
changes in it, was increasingly painful.
Purely mechanical split into 8 files in the same package. No
signatures, behaviors, or types were modified. knowledgeService
struct, NewKnowledgeService, and package-level errors remain in
knowledge.go unchanged. Per-file imports were trimmed by goimports.
knowledge.go 497 struct + constructor + errs + CRUD reads + Search
knowledge_create.go 1144 CreateFrom{File,URL,Passage,Manual} + helpers
knowledge_delete.go 686 Delete + wiki cleanup + ProcessKnowledgeListDelete
knowledge_process.go 2407 ProcessDocument/Summary/Question + Reparse + UpdateImageInfo
knowledge_clone_move.go 1049 Clone* + Move* + progress persistence
knowledge_faq.go 1932 FAQ CRUD + Search + Export
knowledge_faq_import.go 1939 FAQ import/validate/index + runningFAQImportInfo
knowledge_util.go 352 file-type/url/hash + VLM+Storage config + resolveFileService
Verification: go build, go vet, gofmt on the 8 files, and
golangci-lint all clean (22 issues post-split matches 22 pre-split
in knowledge.go — no new issues introduced). The pre-existing
unused warnings on getVLMConfig / buildStorageConfig were left as-is
rather than removing dead code beyond the scope of this task.
Resolves the issues surfaced by a self-review of the adaptive chunking
work. Tests covering each regression were added first to prove the
fix.
Critical
- Heading splitter no longer prepends breadcrumb to Content. The
breadcrumb now lives on Chunk.ContextHeader; EmbeddingContent()
combines header + content for the embedder. This restores the
End-Start == len(Content) invariant relied on by knowledge.go's
document-reconstruction path (summary generation, UI highlight) and
also eliminates the duplicate-heading regression where the section
heading appeared twice in a chunk's body.
- splitBySeparators is now genuinely recursive: when one separator
leaves a piece > ChunkSize, the next-priority separator is applied
inside that piece. Mirrors the Python reference. Without this, a
document with one paragraph break followed by a long run of
newline-separated lines could emit a single ~1950-rune chunk for
ChunkSize=300.
Medium
- SplitParentChild forces children onto the recursive tier, skipping
the per-parent profile pass (was N extra O(N) scans).
- Heuristic splitter snaps overlap start to the nearest semantic
boundary or newline instead of slicing mid-line / mid-word.
- Strategy.Split returns the legacy tier's output directly when the
whole chain rejects rather than running SplitText a second time.
Low
- ensureDefaults caps ChunkOverlap at ChunkSize/2 so pathological
configs (TokenLimit clamping ChunkSize tiny while Overlap stays
high) cannot produce ~97% overlap chunks.
- DefaultChunkOverlap docstring spells out the migration story for
KBs that stored ChunkOverlap=0.
ContextHeader plumbing
- chunker.Chunk and types.ParsedChunk gain ContextHeader plus an
EmbeddingContent() helper.
- types.Chunk gets a transient (gorm:"-") ContextHeader field so the
knowledge service can build embedding inputs as
titlePrefix + chunk.EmbeddingContent() without a DB schema change.
https://claude.ai/code/session_01XADhx6mtu2ZYW3DE9Lun6k
Reciprocal Rank Fusion previously hard-coded k=60 and weights 0.7/0.3.
Both are now sourced from RetrievalConfig (RRFK, RRFVectorWeight,
RRFKeywordWeight) with the same numbers as defaults via three new
GetEffective accessors. fuseOrDeduplicate / fuseWithRRF accept the
config; HybridSearch reads it from tenantInfo and forwards.
Lets corpus owners tune fusion behavior — large corpora benefit from
larger k, keyword-heavy domains from a higher keyword weight, etc. —
without changing application code.
https://claude.ai/code/session_01XADhx6mtu2ZYW3DE9Lun6k
batchEmbedWithBackoff replaces the previous fixed 100ms × 5 retry loop
with proper exponential backoff (200/400/800/1600/3200 ms) and respects
context cancellation between attempts. sanitizeForEmbedding caps any
single embedding input at 20k runes — well above realistic chunk
budgets, so it only triggers on pathologically large inputs and logs
a warning when it does.
https://claude.ai/code/session_01XADhx6mtu2ZYW3DE9Lun6k
The Go splitter previously combined all separators into one regex, losing
the priority-based recursive behavior of the Python reference. This pass
walks the separator list in order: the first separator that yields >1
piece wins, mirroring docreader/splitter/splitter.py. Larger semantic
boundaries (\n\n) now win over smaller ones (\n).
Additional cross-cutting changes:
- ChunkingConfig gains Strategy / TokenLimit / Languages fields so the
per-KB config can pick a tier and pin embedding token budgets.
- buildSplitterConfig in the knowledge service uses the new
chunker.DefaultChunkSize / DefaultChunkOverlap constants (overlap
consolidated to 80 ≈ 15%) instead of the previous hardcoded 50.
- ensureDefaults clamps ChunkSize to a per-language character budget
derived from TokenLimit so chunks stay safe under embedding model
token caps. Chinese (~1.7 chars/tok) gets a tighter clamp than
English (~4 chars/tok).
https://claude.ai/code/session_01XADhx6mtu2ZYW3DE9Lun6k
Adds the building blocks for tiered chunking strategies — multilingual
patterns, token approximation, document profiling, validation, and a
strategy resolver that today still falls through to the legacy
SplitText. Default Strategy is empty/legacy so existing knowledge bases
continue to chunk identically.
- tokens.go: char/token approximation per language (en/de/zh/mixed)
and a coarse DetectLanguage heuristic
- patterns.go: source-of-truth regex bibliothek for headings, numbered
sections, multilingual chapter markers, visual separators, etc.
- profiler.go: single-pass document scan + SelectStrategy chain
- validator.go: rejects obviously broken chunk-sets to drive fallbacks
- strategy.go: public Split / SplitParentChild entry points; resolves
StrategyAuto via the profiler, anything else maps to a fixed tier
- splitter.go: SplitterConfig gains Strategy / TokenLimit / Languages
fields and DefaultChunkSize / DefaultChunkOverlap constants (overlap
raised to 80 ≈ 15%)
- knowledge.go: switches both call sites to chunker.Split /
SplitParentChild — behavior unchanged for empty Strategy
https://claude.ai/code/session_01XADhx6mtu2ZYW3DE9Lun6k
knowledgeService.buildStorageConfig had hardcoded switches in two places that
only handled local/minio/cos. Knowledge bases configured for tos, s3, oss, or
ks3 reached the docreader with a DocParserStorageConfig containing only the
Provider tag — bucket, endpoint, and credentials were silently dropped, and
extraction stalled or fetched from the wrong location.
- Tenant-merge switch: add tos/s3/oss/ks3 cases mapping
StorageEngineConfig.{TOS,S3,OSS,KS3} fields onto DocParserStorageConfig.
Same shape as the existing minio/cos branches; AccessKey -> AccessKeyID,
SecretKey -> SecretAccessKey, no AppID.
- Legacy compat switch: list tos/s3/oss/ks3 alongside local with hasKBFull=false
so fall-through to the tenant-merge path is intentional rather than an
unrecognised provider sliding past the switch. The legacy `cos_config` column
predates these providers and they should never resolve via that path.
- Doc drift: types/custom_agent.go ImageStorageProvider comment and
handler/system.go StorageEngineStatusItem.Name comment now list all 7
providers (matched against types.StorageEngineConfig and ParseProviderScheme).
Tests (knowledge_storage_config_test.go, +268 LOC, 17 cases):
- TestBuildStorageConfig_TenantMergeAllProviders pins per-provider field
population for all 7 providers; the tos/s3/oss/ks3 sub-tests are the
regression guard for #1117.
- TestBuildStorageConfig_LegacyPathOnlyForCOSAndMinIO pins that legacy
StorageConfig data is only consumed for cos/minio and skipped for the
others (intentional fall-through).
- TestBuildStorageConfig_NoTenantFallsThroughToEmpty pins the no-tenant case.
Out of scope (called out in the linked issue):
- The systemic refactor (single source-of-truth provider registry) — left as
a follow-up; this PR keeps changes local to the buggy code path.
- initialization.go validateStorageEngineConfig multimodal validation gap —
the InitializationRequest struct does not carry tos/s3/oss/ks3 fields, so
fixing it requires API surface additions; can ship separately.
Closes#1117 (the buildStorageConfig switches; remaining items deferred).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
internal/application/service/knowledge.go had grown to 9883 lines /
149 functions spanning CRUD, document processing, summary/question
generation, clone/move, FAQ (CRUD+import+index+export), wiki cleanup,
multimodal image, and file utilities. Navigating it, or diffing
changes in it, was increasingly painful.
Purely mechanical split into 8 files in the same package. No
signatures, behaviors, or types were modified. knowledgeService
struct, NewKnowledgeService, and package-level errors remain in
knowledge.go unchanged. Per-file imports were trimmed by goimports.
knowledge.go 497 struct + constructor + errs + CRUD reads + Search
knowledge_create.go 1144 CreateFrom{File,URL,Passage,Manual} + helpers
knowledge_delete.go 686 Delete + wiki cleanup + ProcessKnowledgeListDelete
knowledge_process.go 2407 ProcessDocument/Summary/Question + Reparse + UpdateImageInfo
knowledge_clone_move.go 1049 Clone* + Move* + progress persistence
knowledge_faq.go 1932 FAQ CRUD + Search + Export
knowledge_faq_import.go 1939 FAQ import/validate/index + runningFAQImportInfo
knowledge_util.go 352 file-type/url/hash + VLM+Storage config + resolveFileService
Verification: go build, go vet, gofmt on the 8 files, and
golangci-lint all clean (22 issues post-split matches 22 pre-split
in knowledge.go — no new issues introduced). The pre-existing
unused warnings on getVLMConfig / buildStorageConfig were left as-is
rather than removing dead code beyond the scope of this task.
Addresses two pre-merge review notes:
1. Extract the sentinel-returning threshold check from getSummary into
a small package-private helper (checkSufficientSummaryContent) and
add a unit test for it. The new test pins the behavior the caller
in ProcessSummaryGeneration relies on: the sentinel is returned for
empty / image-only / sub-threshold / scanned-PDF-without-OCR inputs,
and nil is returned for legitimate short notes and scanned PDFs
where VLM OCR succeeded.
The downstream "errors.Is(err, errInsufficientSummaryContent) →
SummaryStatusFailed + skip summary chunk" branch is mechanical (no
branching beyond the errors.Is check); covering it end-to-end would
require mocking KnowledgeRepository / ChunkRepository / chat.Chat,
for which there is no precedent in this package.
2. Switch minTextContentRunes from const to var so tests can override
the threshold and so future config plumbing can adjust it at
runtime without a rebuild. The new test exercises the override.
Drive-by comment fix: extractEntitiesAndConceptsNoUpsert still claimed
summary slugs are code-generated from the document title. Since the
previous commit they are derived from the knowledge ID instead.
Pure documentation change, no behavior impact.
Self-review of the prior commit surfaced one critical regression and
several smaller issues. This change addresses them.
Critical: the old stripImageMarkup stripped the WHOLE <image>...</image>
block, but searchutil.EnrichContentWithImageInfo wraps successful VLM OCR
output in exactly that shape — so a scanned PDF where OCR worked was
being rejected by hasSufficientTextContent and routed to the same
"failed" path as the OCR-empty case. The fix splits stripping into:
* <image_original>...</image_original> blocks removed entirely
(they only carry the redundant Markdown ref already stripped above)
* <image>, <images>, <image_caption>, <image_ocr> tags removed at
the tag level, INNER TEXT PRESERVED, so OCR / caption text reaches
the threshold check and the LLM call.
A regression test pins this down with realistic enriched-image XML that
the previous code failed silently on.
Other fixes:
* Threshold lowered from 50 to 10 runes — short legitimate documents
(brief memos, single-line notes) must still pass.
* Summary chunk content no longer prefixes "# Document\n<filename>";
only the LLM-generated summary is indexed, so filenames cannot
leak back into RAG retrieval context.
* Wiki summary slug is now derived from the knowledge ID rather than
the upload filename, eliminating filename-based slugs like
"summary/mx5280-pdf" that downstream wiki-edit prompts can read.
* Citation sourceRef likewise drops the "|<docTitle>" suffix.
* Dead docTitle parameter removed from extractCandidateSlugs,
extractEntitiesAndConceptsNoUpsert, classifyChunkCitations.
* The redundant stripImageMarkup pass in the warning log is gone;
realTextRuneCount caches the count once and uses
utf8.RuneCountInString to avoid the rune-slice allocation.
Wiki-page display titles still flow through DocTitle (filename) into
later wiki-edit / index-rebuild prompts. That is a larger refactor of
the wiki pipeline and is intentionally left for a separate PR.
Scanned PDFs without a successful OCR pass (e.g. when no VLM is configured
on the KB, or when the VLM returned a "no text content" reply that the
sanitizer discarded) reach the summary and wiki-ingest pipelines with
nothing but markdown image references for content. The downstream LLM
prompts used to receive the filename, file type, and title alongside that
empty content, and the generate_summary template explicitly forbade the
model from refusing — so the model would invent a topic from the filename
alone, producing e.g. a "Canon scanner manual" summary for an MX5280.pdf
that is actually a scanned legal letter.
Two changes shut this down:
1. Drop filename / title / file-type from every LLM prompt input. The
wiki summary, knowledge extract, candidate slug, and chunk citation
prompts now see only the document content. The knowledge summary
path no longer prepends the "Document Type: X / File Name: Y" intro.
2. Add a content-sufficiency guard in front of the LLM call. A new
helper strips markdown image references and HTML image tags and
refuses to invoke the model when fewer than 50 real-text runes
remain. getSummary returns a sentinel error in that case and the
caller marks the knowledge as SummaryStatusFailed instead of
surfacing the first chunk's bare image reference. Wiki ingest skips
LLM extraction with a clear log message.
The generate_summary.yaml prompt is also updated: the "NEVER output
refusal phrases" clause is removed and replaced with an explicit
empty-content rule that returns a deterministic placeholder line.
Real OCR (Tesseract / PaddleOCR) and filename-pattern sanitisation are
intentionally left for separate PRs.
Add Kingsoft Cloud KS3 as a supported storage provider, including
backend service implementation, storage config UI, and provider
scheme recognition throughout the codebase.
AgentFinalAnswer may carry Done:true more than once (e.g. finish chunk
plus EOF). Guard AddEpisode with sync.Once so memory persists once per
turn, consistent with qa.go completion handling.
Co-authored-by: Cursor <cursoragent@cursor.com>
- Apply the session-list additions to the SQLite init migration (Lite build)
and make QueryPaged dialect-aware: LOWER(..) LIKE on SQLite, ILIKE on
Postgres, drop NULLS LAST on SQLite. Escape LIKE wildcards in the keyword
via the existing escapeLikeKeyword helper.
- SetPinned now returns rowsAffected so the handler can respond 404 on
unknown / unauthorized session IDs instead of a misleading 200.
- GET /sessions always returns the enriched shape (pin state + IM origin
fields) so the frontend never needs a second roundtrip; the dual-path
legacy branch in the handler is gone.
- Register pin routes with the wildcard name that matches each verb's
existing radix tree (POST :session_id, DELETE :id) and accept either
param name in the handler; avoids gin's "wildcard conflicts" panic.
- Drop the redundant [platform] prefix from IM session titles now that
the list renders a platform icon alongside the title; add unit tests
for shortID / buildUserSessionTitle / buildThreadSessionTitle.
- Frontend: remove the submenu search input and its i18n/keyword wiring
(search lives elsewhere in the app); pin icon uses the TDesign `pin`
glyph and inherits color so the active session turns green; optimistic
pin moves the item to the top of the list so it shows up at the top
of the Pinned group; IM text badge replaced by the platform SVG icons
from assets/img/im, desaturated by default and full-color on hover
or when the session is active.
Sessions today are a flat list ordered by updated_at. Two gaps showed up in
practice:
- Users cannot find specific chats as the list grows beyond a screen.
- IM-created sessions (WeCom/Feishu/Slack/...) are indistinguishable:
every title was "IM-<platform>" or "IM-<platform>-<username>" and the
list API hid the underlying im_channel_sessions mapping, so admins had
no way to tell which Feishu group a session came from.
Backend
- Migration 000039 adds sessions.user_id (owner), is_pinned, pinned_at
plus a composite index for the list query. Existing rows keep user_id
NULL and stay visible at the tenant level for backward compatibility.
- CreateSession now writes the caller's user_id from auth context.
- GET /sessions accepts keyword / source / agent_id. When any filter is
set, the response switches to enriched items that LEFT JOIN
im_channel_sessions and expose im_platform / im_chat_id / im_thread_id
/ im_user_id / im_agent_id / im_channel_id. No filters => legacy shape,
existing clients unaffected.
- Ordering: is_pinned DESC, pinned_at DESC NULLS LAST, updated_at DESC.
- POST/DELETE /sessions/:id/pin for user-scoped pin/unpin.
- IM session titles: "[platform] <user|chat|thread>" with short ID
suffixes so group/DM/thread sessions are visually distinct without
needing a round-trip to fetch a display name from the IM adapter.
Frontend
- Search input debounced at 300ms drives the keyword filter.
- Pinned chats render in a dedicated group above the time-based groups,
with a pin icon and a pin/unpin entry in the per-chat dropdown.
- IM chats get a short [platform] badge in the list.
- Pin toggle is optimistic and guards against double-clicks.
- zh/en/ko/ru i18n keys added for the new strings.
hjz