Two backend fixes for the v0.7.0 regression where a deleted document keeps
showing in the list ('delete success' toast) and a follow-up reparse then
reports the file is missing:
1. List filtering: applyKnowledgeListFilter now hides rows in the transient
'deleting' state from the default document list. The async delete pipeline
marks a row 'deleting' before tearing down its resources, so without this
the entry lingers as a normal doc until it is physically gone. Rows whose
delete task exhausts retries are flipped to 'failed' by the dead-letter
callback and stay visible so the failure remains actionable; an explicit
parse_status=deleting filter can still surface in-flight rows.
2. Delete ordering: DeleteKnowledge / DeleteKnowledgeList now delete the DB
row FIRST and defer physical file (+ extracted image) cleanup until after
the row is gone. Previously file removal ran concurrently with the
index/chunk/graph cleanup; when one of those failed the row was kept but
the file could already be deleted, producing a 'file missing but row
present' zombie that could neither be reparsed nor cleanly re-deleted.
Orphaning a file after the row is gone is the tolerable failure mode.
Adds a repository regression test covering the default/explicit deleting
visibility rules.
Store authorization timestamps as timestamptz and normalize Unix expiry
inputs to UTC so tenant/platform API keys expire at the intended instant
even when the app process timezone differs from the PostgreSQL session.
Cloned images were unreachable via GET /knowledge-bases/:id/files:
copyOwnedObject used CopyFile, which writes to the knowledge-scoped upload
layout ({tenant}/{knowledgeID}/...), but the KB proxy only serves objects
under {tenant}/exports/ (ValidateKBScopedStoragePath). Read the source
bytes and re-save via SaveBytes instead, mirroring how images are
persisted during ingestion, so copies land in the servable exports
namespace and remain independent of the source object.
For document knowledge, extracted images live in independent image_ocr/
image_caption child chunks that carry image_info, while the parent text
chunk embeds the  reference with an empty image_info. Rewriting
content per-chunk therefore missed the parent text chunk, leaving it
pointing at soon-to-be-deleted source storage.
Move content URL rewriting to a final pass over all cloned chunks, run
once urlCache holds the complete old->new mapping. FAQ chunks are
self-contained, so they are rewritten inline from the same cache.
When cloning knowledge chunks, image_info was deep-copied to new storage
objects but Content still embedded the source markdown image URLs. After
the source knowledge is deleted, those in-content references break while
image_info points at the new copies. Rewrite Content alongside image_info
so both stay consistent.
SendReply now sets param.formatType=markdown by default since WeKnora
replies are authored in Markdown. Callers can override via
reply.Extra["yunzhijia_format_type"] (empty string disables the param).
- Added AminusB method to the knowledgeRepository, which returns IDs of knowledge in A that have no counterpart in B, comparing by file_hash as a multiset.
- This implementation addresses issues with handling duplicate entries and NULL file_hash values, ensuring accurate synchronization between knowledge bases.
- Updated the KnowledgeRepository interface to reflect the new functionality and improved documentation for clarity.
Introduce platform API keys with capability-based access to tenant management,
system settings, runtime queues, and audit logs, plus admin UI for key lifecycle
and system audit log viewing.
Large Feishu wikis stopped syncing after the first batch (Tencent/WeKnora#2136:
5000+ docs, only ~150 ingested; #1262: a child-list failure aborted the run).
Three compounding causes, fixed in layers:
1. Rate-limit / transient failures had no retry. Feishu's wiki/drive/export
APIs are aggressively rate limited; a thousand-doc sync issues tens of
thousands of calls, and a single 429/5xx burst failed whole swathes of docs.
Add exponential backoff honouring Retry-After to doRequest (JSON) and
downloadRawBytes (files): 429 up to 3x, 5xx once, transport errors back off.
2. The batch path fetched the entire tree into memory before ingesting, so a
sync that hit the 2h task timeout lost all progress and restarted from
scratch, and a transient per-document export failure advanced the cursor and
skipped that doc permanently. Add an optional StreamingConnector interface,
implemented for Feishu: FetchStream interleaves fetch -> ingest -> checkpoint,
bounding memory to one item and persisting a per-node edit-time cursor at
page boundaries. The service dispatches to it via type assertion; connectors
without it fall back to FetchAll / FetchIncremental unchanged (non-breaking).
A resumed/incremental run skips nodes already recorded at their current edit
time (converges), a failed fetch does not advance the cursor (retried next
run instead of dropped), and partial listing failures emit error items and
carry prior edit times forward instead of aborting the sync.
3. Failures and skips were invisible. Unsupported node types (mindnote/slides/…)
were skipped with no item, error or log ("13 docs synced only 3"). FetchAll
now tallies discovered/fetched/failed/skipped per resource with a per-type
skip breakdown and a progress line every 100 nodes. The backend already
persisted per-item errors in SyncLog.result.errors but the frontend type
omitted the field; declare `result` on SyncLog and render up to 50 failed
documents in the sync-log drawer (#1262). i18n for zh-CN/en-US/ko-KR/ru-RU.
result.errors is a jsonb column shipped in every list response, so it is
capped at 100 samples server-side (recordSyncError); the exact failure total
comes from the bounded items_failed counter, which the drawer uses for its
count and "N more" line so a sync that fails thousands of docs stays bounded.
Each sample is a structured, localisable SyncItemError instead of a raw
"title: ...api error: status=500 body={json}" dump that repeated the title
and leaked status/body/log_id. The connector (feishuFailure) classifies the
error into a stable i18n code (+ the feishu code as a param); the service
builds {title, code, params, message}; the drawer localises via vue-i18n to
the viewer's language (zh/en/ko/ru), falling back to message. A custom
UnmarshalJSON keeps old string-form logs readable. The banner is localised
from items_failed too. Raw errors stay in the server logs. Follows the
two-tier norm of Airbyte/Fivetran/Onyx: human, actionable, localised
summaries in the UI; raw detail in logs. Transient errors say "will retry"
(the cursor is retained); auth/permission errors point at the fix instead.
Tests: client_retry_test (backoff), tally_test (accounting), connector_stream /
datasource_stream (streaming dispatch + checkpoint), connector_convergence
(timeout + transient-failure resume converges to full coverage; proven to catch
#2136 via bug injection), datasource_result_cap (error-sample cap + structured
localisable SyncItemError, no raw body), connector_error_reason (error→i18n-code
classification), sync_item_error (legacy-string + structured back-compat) — all
in the normal suite. connector_realapi / connector_seed are an opt-in real-Feishu
harness gated by the feishu_integration build tag AND env vars, never run in CI.
The generic OpenAI-compatible reranker hardcoded `truncate_prompt_tokens:
511` on every /rerank request. Providers that honor this vLLM-specific
field (e.g. SiliconFlow serving Qwen3-Reranker) keep only the last 511
tokens of the templated rerank prompt, cutting the instruction and query
off any document longer than ~511 tokens. Every relevance score then
collapses to near zero (~0.0067 instead of ~0.998), falls below the score
threshold, and knowledge-base search returns empty results. A direct curl
without the field returns the correct scores.
WeKnora already removed this exact hardcoded parameter from the embedding
path in #960; the rerank path was never fixed.
- Add `omitempty` to `truncate_prompt_tokens` and `additional_data` so
empty optional fields are not sent.
- Make truncation opt-in via `extra_config.truncate_prompt_tokens`
(validated > 0 at construction), preserving an escape hatch for
self-hosted vLLM deployments with small-context rerankers.
- Add regression tests reproducing the collapsed-score symptom and the
opt-in path.
Fixes#2143
Multiple GORM tag options must be separated by ';', not ','. The comma was
parsed as part of the type value, producing malformed DDL that causes
AutoMigrate(&Message{}) to fail on SQLite with a syntax error.
Restrict channel-managed sessions (API keys, IM, embed) to tenant admin
in the Web console for list/read paths, hide channel sidebar buckets from
non-admins, and add CountSessionsBySource for aggregate stats without
exposing session rows to viewers.
Split oversized candidate sets into concurrent batches instead of failing when retrieval exceeds the VikingDB rerank per-request cap, and align ko/ru locale hints with zh-CN.
Let tenant-wide retrieve keys serve tenant-bounded /files paths while
denying KB-restricted keys, declare knowledge batch write routes for
ingest-capable keys, and short-circuit ownership checks for API-key
principals already authorized by the gate and KB allow-list.
Add GetOwnedSession method to enforce strict ownership checks for session mutations, ensuring tenant admins can read but not modify API-key sessions. Update relevant handlers to utilize this method for operations that alter session data, maintaining security and integrity of API-key sessions.
Let tenant admins list and open API-key sessions from the Web console while
keeping per-key isolation for everyone else. Pin the session source filter so
it no longer jumps when switching buckets, and resolve session origin on read
so deep links stay in sync with the active folder after refresh.
Use bounded detached contexts for wiki ingest tail cleanup so pending row
settlement, span closure, finalize enqueueing, and follow-up scheduling can
finish even after the asynq task context is cancelled.
Return settlement errors from pending row trimming and failed-op requeueing,
and only mark claims settled after rows are deleted, released, or
dead-lettered successfully. Requeue documents whose reduce updates were not
applied after context cancellation to avoid silently dropping wiki work.
Add unit coverage for detached cleanup contexts and settlement error handling.
Finish-time metadata was discarded and implicit root traces opened by
StartSpan/StartGeneration were never ended, leaving orphaned parents in
Langfuse. Also stop mutating the global OTel text-map propagator.
Migrate WeKnora's Langfuse tracing from the legacy /api/public/ingestion
batch API to the OpenTelemetry Go SDK emitting OTLP/HTTP to
/api/public/otel/v1/traces — the Langfuse v3+ standard ingestion protocol.
Why: Langfuse v3 deprecated the legacy ingestion API in favor of OTLP, and
OTel-only backends (e.g. the LiteFuse Doris fork) dropped it entirely, so the
legacy client was incompatible. Migrating to the OTel SDK also adds W3C
traceparent cross-service propagation: an upstream caller's trace id is
inherited so its run and the WeKnora call land under one trace in the UI.
What:
- Manager owns an OTel TracerProvider backed by an otlptracehttp exporter
(Basic auth + the x-langfuse-ingestion-version gate header). Spans are
buffered by BatchSpanProcessor and exported complete on End, so a long
trace no longer fragments across flush batches.
- *Trace/*Span/*Generation wrap OTel spans; the public API (StartTrace/
Span/Generation, handles, InjectTracing, asynq/gin middleware, Config,
TokenUsage) is unchanged, so callers (agent/chat/handler/scheduler/asynqdl)
need no edits.
- W3C traceparent: GinMiddleware extracts traceparent from incoming requests;
AsynqMiddleware + InjectTracing propagate it across the asynq payload.
types.TracingContext gains a LangfuseTraceparent field.
- logger.CloneContext preserves the active OTel span across context rebuilds
so child spans (agent engine, summary generation) stitch to the HTTP trace
instead of orphaning into separate traces.
- StartSpan/StartGeneration re-establish the parent from the *Trace handle
when the ctx lost the active OTel span, covering background/derived contexts.
Compatibility: works with Langfuse v3+ (cloud / self-hosted / the bundled
docker-compose profile) and LiteFuse. Drops Langfuse v2 (legacy-only) support.
Uses the OTel Go SDK already (indirectly) in go.mod, promoting
otlptracehttp to direct; no third-party langfuse-go library.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Chat attachment parser rules belong on agents; drop the duplicate workspace settings section while preserving legacy tenant rules when engine config is updated without that field.
Decouple the references SSE event from the per-agent citation output toggle
so the retrieval-results drawer still works, and align follow-up suggestion
card layout with the chat column.
Lark and Feishu are the same product deployed on two isolated clouds
(open.larksuite.com and open.feishu.cn). WeKnora touches both from two
places — the IM bot and the Wiki data-source connector — and this brings
Lark to parity across both. Neither API surface differs, so rather than
duplicate either implementation, each is parametrized by a Region and
registered twice.
IM (internal/im/feishu):
Every API call routes through a.api(), which prefixes the region's
host, and the WebSocket client passes larkws.WithDomain — the SDK
defaults to open.feishu.cn, so Lark long connections would otherwise
fail to authenticate.
Data source (internal/datasource/connector/feishu):
Lark was half-supported: the client honoured a base_url override and a
LarkBaseURL constant existed, but the connector registered only as
"feishu" and the UI never exposed base_url, so Lark wikis could not be
configured at all. Lark is now its own connector type. base_url is kept
as an explicit override so data sources that pointed a "feishu"
connector at open.larksuite.com keep working.
Bugs found and fixed along the way:
- Wiki resource links were hardcoded to https://feishu.cn/wiki/{token} in
four places, so a Lark wiki linked into the wrong cloud entirely.
- Bot identity is now platform-prefixed (feishu:cli_x vs lark:cli_x).
BotIdentity carries a unique index, so without this a Lark channel
would be rejected as a duplicate of an unrelated Feishu bot sharing an
app_id.
- The card image cache is now keyed by app_id. image_keys are issued per
app on one cloud; a Feishu key handed to a Lark app fails the card
update with code=200570.
- The datasource doc claimed Lark "adapts automatically" — it could not,
since the UI had no way to reach base_url.
Permissions are where the clouds genuinely diverge. Identical APIs do not
mean identical permission catalogues: aily:file:read, aily:file:write and
corehr:file:download exist on Feishu but not on Lark, so pasting Feishu's
scope JSON into Lark fails to import. That JSON is also over-broad (it
covers products WeKnora never calls) and simultaneously incomplete (it
omits the drive:*/docx:* scopes the connector needs). Lark therefore gets
its own list, derived from the 17 endpoints the code actually calls and
checked identifier-by-identifier against Lark's own API reference, plus a
full Feishu→Lark mapping table marking each scope verified, unverified,
or absent. Both features use tenant_access_token only — no user_access_
token exists anywhere in the repo — so the user dimension stays empty.
Known limitation: Region.ThinkingText makes Lark's initial card
placeholder English, but once streaming starts the think-block headers
come from the shared MarkdownThinkStyle and are Chinese on every platform
(Lark, Slack, Telegram, Mattermost alike). Fixing that means threading a
locale through FormatIMDisplayContent, which has no platform context — a
cross-platform change, not a Lark one. Documented as such.
The channel-creation 400 message is now derived from validIMPlatforms so
the list cannot drift as platforms are added.
Tests: region invariants and per-region URL construction on both sides,
the Feishu/Lark bot-identity split, platform-to-channel mapping, base_url
override precedence, an end-to-end SendReply against a fake Open Platform
(token fetch, reply, thread-unsupported fallback, hard-error surfacing),
and a ListResources drive asserting wiki links never leak the other
cloud's host.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Update the SuggestedQuestionsRequest to include TagScopes for better filtering of knowledge items based on their parent knowledge bases. Modify related API documentation and frontend interfaces to accommodate the new structure. Ensure backend services handle the new tag scope logic correctly, including tests for the updated functionality.
Move citation protocol injection to the system layer so custom prompts no
longer own citation syntax. Agents can disable inline citations while source
aliases still encode and decode tool arguments correctly.
Replace raw chunk/document/KB IDs in LLM context with short cN/dN/bN/wN
aliases, decode tool arguments on the way in, and expand compact <ref/>
citations to the public <kb>/<web> format before user-facing output.
Implement a maximum limit of 5 for the total number of images and attachments that can be uploaded in a chat session. Introduce a warning message for users when the limit is exceeded. Additionally, add functionality to clear pending uploads when switching sessions to prevent errors related to stale document IDs. Update localization files to include new warning messages in English, Korean, Russian, and Chinese.
Add functionality to normalize temporary attachment IDs, ensuring deduplication and enforcing a maximum limit of 5 attachments per message. Update related handlers and tests to support this new behavior, enhancing the robustness of attachment management in chat sessions.
Persist resolved temporary-attachment content onto the stored user message
so later turns can replay it in history. Close the retrieval progress spinner
on early pipeline exits, and reshape model-fallback messages with a proper
system role plus trailing user question.
Update the classification logic for follow-up questions to include reasoning about previously described images or documents. Ensure that follow-up queries referencing earlier attachments are correctly identified and processed without triggering unnecessary knowledge base searches. This change improves the model's ability to maintain context and respond accurately to user inquiries related to past interactions.
Let users upload documents in chat without ingesting them into a knowledge base: files are parsed in the background, referenced by ID on send, and expire after a configurable TTL.
Tighten model prompt so retrieval questions stay topic-bound and self-contained,
reserve knowledge slots in hybrid mode, and carry suggestion KB IDs into the next send.
When retrieved context or tool results contain Markdown images, dynamically
append output requirements to system prompts and final-answer instructions.
Also normalize fullwidth image parentheses in the frontend renderer so
localized model output still renders correctly.
Chat and agent tool context now emits answer-ready Markdown images instead
of internal <image> XML so copied content renders safely. The frontend also
normalizes legacy XML blocks in persisted conversations during streaming.
Improve follow-up question quality by feeding the model structured context
(evidence, capabilities, complete turns) instead of raw RAG content, and
rank knowledge candidates by current-topic relevance.
Refactor tests and registry to use the new 'res://' syntax for resource aliases, ensuring consistency across the codebase. This change addresses alias handling in various test cases, including stream processing and registry encoding, to prevent data loss and improve resource reference management.
Add functionality to detect unresolvable resource aliases emitted by the model during chat completion and streaming. Introduce logging for these orphan aliases to aid in debugging and improve the handling of resource references. Update the stream processing to ensure held aliases are flushed upon cancellation, preventing data loss. Include tests to verify the correct reporting of orphan aliases.
Introduce resource://<handle> as the stable application identity for stored
objects, decoupling API responses, persisted rich text, and LLM prompts from
physical storage paths. Adds a resource catalog service (register/resolve/bind,
revocable expiring access grants), a request-local alias registry that compacts
resource references while in the LLM context and restores them (including across
streaming chunks), plus router, IM, chat-pipeline, and multimodal integration.
Includes schema migrations (resources, resource_bindings, resource_access_grants)
for sqlite, and hardens URL/output security handling.
Implement a comprehensive API for managing storage backends, including endpoints for listing, creating, updating, and deleting storage instances. Enhance the UI to support multi-instance configurations, allowing users to manage multiple storage types effectively. Update localization files for new storage backend features and improve error handling for connectivity tests. This update also includes detailed documentation for the storage backend API.
Enhance error messaging for storage connectivity issues by sanitizing raw errors to prevent leaking sensitive information. Introduce a new utility function for consistent error handling across storage backend operations. Additionally, implement logic to support multi-instance storage backends, allowing for more flexible workspace configurations and improved reporting of active storage providers.
Introduce tenant-scoped storage backend registry so workspaces can register
multiple storage instances and bind each knowledge base to a specific backend,
with migration from legacy tenant-level storage config and new settings UI.
Extract storage allowlist into internal/storageallowlist so KB creation
uses FirstAllowed when tenant default_provider is empty or disallowed.
Share pickUsableStorageProvider via editorResources for consistent UI
selection, and add regression tests for OBS status and allow-list defaults.
Update the manage_kbs capability to cover the full lifecycle of knowledge bases, including create, copy, duplicate, update, and delete operations. Adjust related API routes and tests to ensure proper access control and policy requirements for knowledge base management.