Two follow-up improvements on top of renderDocxFromFile so the docx
pipeline can handle real long-form deliverables instead of just
prose-only memos.
Image embedding (P1).
MarkdownDocxRenderer now recognizes single-line  markdown
and embeds the referenced file via POI's XWPFRun.addPicture():
- PNG / JPG / GIF / BMP read straight from disk
- SVG rasterized via Apache Batik (PNGTranscoder, target width 1400px)
before embedding — OOXML stores raster images, so any vector source
needs conversion. Batik runs in-JVM, no rsvg-convert / cairo on host.
- Pictures are pinned to roughly the printable page width (≈ 5.77 in
for A4 minus default 1800-twip margins) and given a 4:3 height
fallback. Mixing images inline with other paragraph text is not
supported by design — the markdown subset assumes one image per
block paragraph. Inline images would require splitting paragraphs
across runs with explicit positioning, well beyond what this
renderer covers.
- Failure modes (missing file, unsupported format, Batik blowing up)
emit an italicised "[image: alt — reason]" placeholder so the rest
of the document still renders; the agent can read its own log to
see why the picture didn't make it.
- Adds two transitive deps via pom: batik-transcoder + batik-codec at
1.18, ~10 MB combined. Worth it given the alternative is shelling
out to system tooling.
Multi-file render (P2-lite).
New tool renderDocxFromFiles(List<String> filePaths, filename, pageSize)
reads several markdown files in order and renders one combined docx.
Lets the agent split a 30-page proposal into cover.md / ch1.md /
ch2.md / appendix.md and produce a single deliverable in one tool
call. Each path goes through WorkspacePathGuard.validatePath; any
empty or unreadable file aborts with a typed error so the agent
fixes its file list before retrying. Files are joined with a blank
line — no separator markup is injected, headings carry over cleanly.
I deliberately did NOT build the heavier mutable-docx state
("appendDocxChapter / finalizeDocx") flavor of P2: the multi-file
form covers the same workflow with no per-conversation state to
clean up, and the agent can iterate by rewriting the chapter file
and re-running the tool. Stateful append can come later if a
streaming use case actually shows up.
renderDocx and renderDocxFromFile @Tool descriptions updated to point
the agent at renderDocxFromFile for >5 KB markdown and to advertise
the new image-embedding capability.
renderDocx requires the markdown body to flow through the LLM as a
tool argument. For an 80 KB project proposal that's ≈ 20 K tokens of
streaming output spent just to repeat back content the model already
wrote to disk a turn earlier — multi-minute generation, real money.
renderDocxFromFile takes a file path instead. The agent uses
write_file / edit_file to assemble the markdown locally, then calls
this tool with just the path. JVM reads the file in one IO syscall
and feeds it to the existing MarkdownDocxRenderer. Token cost drops
from ≈ 20 K to ≈ 50 (the path string).
Behavior:
- Path resolution honors WorkspacePathGuard, same boundary as
read_file / write_file. No path traversal.
- UTF-8 read; rejects empty / missing / non-regular paths with
typed error messages so the agent can recover.
- Output cached in GeneratedFileCache and returned as a relative
/api/v1/files/generated/{id} link, with the same anti-host-
hallucination instruction renderDocx already carries.
- Same supported markdown subset (headings, bold, lists, tables).
Image references () still render as raw text — full
image embedding (P1) and SVG → PNG conversion (also P1) need
Apache Batik plus image-rendering plumbing in MarkdownDocxRenderer
and is tracked separately. Chapter-mode merge (P2) likewise needs
its own plumbing.
The @Tool description tells the agent to prefer this path when
markdown exceeds ~5 KB and shows the full write_file →
renderDocxFromFile workflow inline.
Three small but high-impact fixes that all surfaced together while
verifying the long-form generation flow.
1. ChatConsole onBeforeUnmount no longer kills the backend turn.
Previously, switching tabs / route navigation / any cause that
unmounted the chat view called stopChatGeneration(), which POSTs
/chat/{cid}/stop and aborts the in-flight LLM call. The user
reported a turn dying mid-generation just from switching pages.
Replaced with resetForNewConversation() — front-end SSE disconnect
only, no /stop. Backend keeps running; pollActivity / status probe
reconnects on return. Aligns with the existing comment in
selectConversation: "let A's backend agent run continue running."
2. Agent max_iterations raised 25 → 100 with a hard ceiling.
The previous 25-step ceiling caused LimitExceededNode to fire on
substantive multi-tool tasks (document generation + image conversion
+ retry loops). 100 matches QwenPaw's _MAX_MAX_ITERATIONS upper
bound. New plumbing:
- BaseAgent.MAX_ITERATIONS_HARD_CEILING = 100 public constant
- BaseAgent default field 25 → 100 (Java-side fallback)
- AgentGraphBuilder clamps any per-agent DB override to the
ceiling at runtime; if the row holds 200, runtime sees 100 and
a WARN is logged with the original value.
- V47 migration (h2 + mysql) idempotently bumps the three default
seeded agents (1000000001, 1000000002, 1000000003) only if they
still hold the old defaults (25 / 20). User-customized values
are not touched.
- data-en/zh/-mysql-en/-mysql-zh seed files updated to 100 for
fresh installs.
3. DocxRenderTool tells the LLM not to prepend a host to the URL.
DeepSeek and Claude have both been observed wrapping the
/api/v1/files/generated/{id} relative path returned by renderDocx
into an absolute URL with a hallucinated domain (e.g.
https://ai-tools-system.com/...), breaking the download link in
the rendered chat bubble. The tool's return string now appends an
explicit "must use the relative path verbatim, do not add any
https:// or http:// prefix" instruction, which Claude and
DeepSeek both honor.
A bundle of stability fixes that all surfaced together while running
the same long-form generation task across multiple turns. Each one
addresses a distinct way the previous behavior silently dropped
content the user had already seen on screen.
1. Mid-turn narrative persistence (StateGraphReActAgent +
SummarizingNode). Intermediate ReasoningNode rounds and
SummarizingNode broadcast their content_delta directly to the
SSE channel for live display, but the StreamAccumulator only
received the final answer. After refresh the assistant message
showed only tool_call cards with no body text.
StateGraphReActAgent now also forwards STREAMED_CONTENT (already
set per round) as a persistOnly StreamDelta whenever it changes,
so every narrative chunk lands in the accumulator's content
buffer and gets written to mate_message. SummarizingNode now
writes its summary into the same key so summarize narratives
persist too.
2. Follow-up message queue, not dispose (ChatController#interruptStream).
Sending a new message while a turn was running called
requestInterrupt, which dispose()d the active Reactor chain mid
LLM call. That cancelled the in-flight generation, lost partial
tokens, and left the user staring at a half-finished bubble.
The endpoint now uses enqueueMessage in all paths, matching
the "wait for current turn, then run" behavior. The old
requestInterrupt API is kept for any future force-replace UI
but no caller routes to it.
3. Queued user message ordering (ChatStreamTracker.QueuedInput +
ChatController.startQueuedMessage). interruptStream used to save
the queued user message immediately, before the in-flight
assistant message finalized in doOnError. listMessages orders
by create_time ASC, so the queued user message ended up above
the assistant reply it was supposed to follow. QueuedInput now
carries contentParts; persistence is delayed to startQueuedMessage,
which runs only after Asst-N is on disk.
4. JVM shutdown flush (ChatStreamTracker @PreDestroy +
emergencySaveAccumulator). A mvn spring-boot:run restart used to
wipe in-flight turns: SSE emitter timed out, ShutdownHook fired,
HikariPool closed before doOnError could save. ChatStreamTracker
now exposes an emergency-save callback per RunState; ChatController
registers one per stream that snapshots the accumulator and
writes status="interrupted_shutdown". @PreDestroy walks active
runs, invokes the callback, then disposes. Spring's reverse-order
bean teardown keeps ConversationService and Hikari alive long
enough for the save to complete.
5. Observation thresholds for summarize (GraphObservationProperties +
application.yml). The previous total-chars threshold of 12 KB
triggered summarize after one or two RFC reads, costing a 40 to
80 second compaction LLM call per loop. Tuned to: total 200 KB,
single 16 KB, large-result 32 KB, rounds safety net 25. Java
field defaults reverted to the conservative original values so
application.yml stays the source of truth.
6. Frontend thinking segmentation (useChat.ts thinking_delta +
phase). Multi-round ReAct turns merged every reasoning + summarize
round's thinking into one segment, accumulating to 9 KB+ in a
single bubble. thinking_delta now uses findLast(running) so a
tool_call_started or phase transition closes the previous segment
and the next delta opens a fresh one. phase event also closes
running thinking/content segments.
7. Other small things bundled: removed a debug metadata-keys log
that flooded the log file with one line per stream chunk; fixed
three stale tests that didn't compile after earlier constructor
changes (WikiLogServiceTest, WikiOverviewSpliceTest,
WikiProcessingServiceLazyTest); added rfc-066 documenting the
unified message queue + priority refactor as the next logical
step on top of these stabilizations.
Verified end-to-end with multiple full sessions: a four-minute
generation that produced the expected docx and a follow-up enqueue
that ran cleanly after the previous turn naturally completed,
without the old "Disposable unavailable" interrupt path.
Same bug as the prior queue-drop fix in doOnComplete, but in the
sister branch that fires when the agent's reactive stream errors
out (CancellationException from a user stop). The guard
cr.queuedInput() != null && !(isUserStop && !isInterruptFollowup)
mis-classified "user stopped, no interrupt-with-followup, but a
message is in the queue" as an explicit abort and silently dropped
the freshly-typed follow-up.
The frontend's enqueue path never sets interruptType — it just
calls requestStop + offers to messageQueue. Whoever puts a message
in the queue means it; just run it. Aligns with doOnComplete and
the four other queue-launch sites in this controller.
A series of cross-cutting stability fixes that surfaced together
during a long debugging session.
reasoning_content / Claude prefill self-replicating 400:
- ChatController persists typed errors (content starts with '[错误] ')
with status='error', so the failure text stops being re-sent as
multi-turn context — DeepSeek thinking 400 ('reasoning_content
must be passed back') and Claude 400 ('does not support assistant
message prefill') used to recursively re-create themselves every
retry by polluting history.
- BaseAgent.sanitizeForLlm filters status='error' / '[错误] ' prefix
assistant messages from history before LLM dispatch.
- BaseAgent.fetchHistoryMessages defensively drops trailing
AssistantMessages — Claude rejects assistant-tail prompts.
- NodeStreamingChatHelper.dropTrailingAssistant runs the same
defense at every doStreamCall pre-egress, so the in-turn
summarizing→reasoning transition (which leaves an assistant
scaffold at the tail) doesn't trip Claude either.
- AgentGraphBuilder.FallbackPolicy.DEEPSEEK switched (null,true,true)
→ (' ',false,true), aligning with KIMI/OPENAI's tolerant ' '
fallback. The previous 'force explicit 400' design was the
self-replicating loop's prime mover.
narration + tool args truncation:
- ReasoningNode.DEFAULT_MAX_OUTPUT_TOKENS 4096 → 16384. The 4k cap
was decapitating renderDocx tool_call args mid-stream when the
model emitted a long content field on top of thinking content;
the resulting 'invalid JSON' aborted execution silently.
- ReasoningNode appends a hermes-style TOOL_USE_ENFORCEMENT clause
to every system prompt: 'when you say you will perform an action,
call the tool now in the same response — narration is a protocol
violation'. Treats 'now I will generate the docx' (and never
actually calling renderDocx) as a forbidden pattern.
- ToolExecutionExecutor.normalizeToolExecutionError reframes the
JSON-truncated error as actionable instructions: 're-call the
same tool now with shorter content or split into multiple
sequential calls; do NOT describe the result as text'.
side fixes from the same evening:
- ChatController doOnComplete skips completionPublisher.publish
when isError=true, keeping memory extraction off the garbage path.
- ChatController doOnComplete queued-message guard simplified to
'cr.queuedInput() != null', matching the other 4 sites in the
controller. The previous 'isInterruptFollowup || !wasStopped'
guard silently dropped queued messages when the user did
Stop-then-Enqueue (wasStopped=true && interruptType=null), losing
the freshly-typed follow-up message.
- prompts/graph/summarize-system.txt now distinguishes 'single
task' (default; output one cohesive summary) from 'multiple
independent sub-tasks' (use the子任务 N format). Stops the
summarizer from inventing '子任务 1: PRO-027' decomposition for
unitary requests like 'write me a project proposal'.
- docker-compose.yml: pass SEARXNG_BASE_URL into mateclaw-server so the
app can reach the searxng sidecar container out of the box (default
http://searxng:8080).
- SystemSettingService: resolveSearxngBaseUrl() now falls back to the
SEARXNG_BASE_URL env var when no DB value is set, so Docker users no
longer need to configure it manually in the UI.
- V38 migration (h2 + mysql): expand mate_wiki_chunk.content from TEXT
(64KB) to MEDIUMTEXT (16MB) so large Chinese chunks (~30k chars
≈ 90KB UTF-8) no longer overflow.
Five-commit bundle brings the Dream v2 P1 engine layer online, sitting
on top of the lifecycle mediator foundation already merged.
B.1-B.4 · Schema + records
- Flyway V26 (dream_report) + V27 (memory_recall review fields),
both h2 and mysql
- DreamReportEntity + DreamMode + DreamStatus enum + record types
- DreamReportMapper repository layer
B.5-B.8 · Consolidate refactor + focused dream
- MemoryEmergenceService refactored for plug-in dream modes
- MemoryRecallService extended with promoted/rejected review fields
- Focused dream endpoint + prompt template
- MemoryController exposes the review/trigger surface
B.9-B.10 · Monthly archive service
- MemoryArchiveService rolls cold promoted entries into archival rows
and reclaims daily_count storage
- DreamingScheduler runs archive job on its own schedule
B.12-B.14 · Tests
- MemoryArchiveServiceTest
- DreamFlagGuardTest
- DreamV2AcceptanceIT (end-to-end acceptance under feature flag)
Plus a verification script + HTTP e2e kit in the private test/ dir,
used for local staged rollout — not part of the open-source
distribution.
All features stay gated behind the mate.memory.dream.* flags from
Phase 1. Enable per-phase after staging validation.
beforeLlmCall / afterLlmCall / onSessionEnd only logged on failure,
making flag on/off indistinguishable in logs. Add debug lines on the
success path so lifecycle activation is observable.
Wire memory-facing events (turn-started, turn-completed, session-ended,
memory-written) through a single MemoryLifecycleMediator so
MemoryProvider implementations can hook into the agent conversational
flow without spreading side-effects across the runtime.
Ten atomic steps shipped under feat/dream-v2-p1-lifecycle:
- A.1 + A.2: MemoryLifecycleMediator class + TurnContext value object
- A.3: TurnStartedEvent / TurnCompletedEvent domain events
- A.4: MemoryLifecycleEventListener bean for Spring event plumbing
- A.5: MemoryProvider.onMemoryWrite default method (backward compatible)
- A.7: wire the mediator into AgentService at the right hook points
- A.8: LifecycleFlagGuardTest — feature flag must gate every hook
- A.9: MemoryLifecycleMediatorTest — unit coverage per hook
- A.10: LifecycleRecallCountIT — F4 regression across the stack
Feature flags (all default OFF; enable per phase after staging):
- mate.memory.lifecycle-mediator-enabled
- mate.memory.dream.focused-enabled
- mate.memory.dream.archive-enabled
This is Phase 1 foundation only — focused-dream and archive-dream
providers arrive in later phases.
Root cause: processRawMaterial() created a job record at queued stage
but never called jobService.transition() during processing. The job row
stayed at queued forever, so the stage bar never advanced.
Backend (WikiProcessingService):
- Transition job to ROUTING immediately after creation
- Transition to PHASE_A_RUNNING before chunk processing begins
- Transition to COMPLETED/PARTIAL/FAILED at the end based on finalStatus
- Transition to FAILED in the catch block on unhandled exceptions
Backend (WikiProcessingJobService.transition):
- Handle FAILED, PARTIAL terminal stages (set finishedAt + status)
- Handle non-terminal intermediate stages (set status to running)
Frontend (JobStageBar.vue):
- Add stageMapping for backend stages not shown as dots: phase_a_done →
phase_b_running, failed/partial/cancelled → completed position
- Guard stageIndex() against -1 (unknown stages default to all-pending)
- Terminal failure states show red failed dot instead of pulsing active
Two related changes that align buildFallbackChain with how users actually
think about failover.
1) Source = configured providers (was: only providers with fallback_priority > 0)
Earlier the chain was strictly "providers the user explicitly opted in via
fallback_priority > 0". A healthy in-pool provider with priority=0 was
silently excluded — surprising since the pool was supposed to be the source
of truth for "what is usable". After this change:
- Candidates = every configured provider
- Pool gating = same as before (in-pool members only at build time;
runtime walker re-checks)
- Order = agent prefs (PR-3) → fallback_priority asc (>0) →
priority==0 alphabetical
So fallback_priority is now purely an ordering hint, never an exclusion.
2) Per-provider model picker = default OR first-enabled (was: default only)
Previously a provider was skipped if no chat model on it had is_default=true.
That is admin friction with no benefit — every provider had to be visited in
Settings just to mark a default before it could appear in failover. New
pickFallbackModel():
- first try getDefaultModelByProvider — user explicit pick wins
- otherwise take the first enabled chat model on the provider
- skip only if neither exists
User-visible effect on the deployment that surfaced this:
- kimi-code primary fails (401 — real auth issue, separate from this bug)
- Pool short-circuits primary → walker fires
- Walker now sees dashscope (in-pool) AND ollama (in-pool) as candidates,
even though neither has fallback_priority set
- dashscope first enabled qwen model is picked → request succeeds via
dashscope without anyone touching Settings
45 failover-related tests still green (unit-level chain-build behavior is
backward-compatible; only the candidate set and model-selection lookups
changed, both broadening the chain rather than narrowing it).
Two real bugs the user restart surfaced — both turned healthy providers
into HARD-removed false positives.
Bug #1 — URL duplication
OpenAiCompatibleListModelsProbe always concatenated /v1/models, so
providers whose Base URL already includes the version segment got the
wrong URL:
LMStudio http://localhost:1234/v1 → /v1/v1/models → 404
ZhipuAI .../api/paas/v4 → /v4/v1/models → 404
Fix: detect a trailing /vN suffix and append /models instead. Six unit
tests in OpenAiCompatibleListModelsProbeTest lock the rule down.
Bug #2 — 404 false positives
Kimi for Coding API does not expose /v1/models even though chat works
fine, so the probe correctly received a 404 and incorrectly HARD-removed
the provider from the pool. Other vendors will hit the same — listing
is not a universal contract.
Fix: classify HTTP responses semantically.
401 / 403 → HARD remove (real auth failure)
404 / 405 / 410 → fail-open (endpoint missing, server may be alive)
other 4xx / 5xx → fail-open (probe inconclusive — let chat decide)
network errors → fail (unreachable)
This is the same philosophy as ChatGPTOAuthStatusProbe: when we cannot
cheaply confirm health, we do not proactively penalize the provider.
Same logic applied to Anthropic + DashScope probes for consistency.
Net effect on the user deployment after restart:
- kimi-code stays in pool (404 → fail-open) → primary path works again
- lmstudio + zhipu-cn also stay in pool (URL bug fixed)
- dashscope + ollama unchanged (real 200 OK)
Tests: 6 new for resolveModelsPath. The 2 unrelated WikiRawMaterialDedupTest
failures pre-date this commit and live in ba86bea.
Root cause: addFile()/addText() hash dedup only matched rows with
status=completed, so the same file uploaded while in partial/pending/
processing/failed status would create a duplicate row.
Fix:
- Remove .eq(processingStatus, "completed") from dedup queries — match
any non-deleted row with the same content hash in the KB
- On dedup hit: completed/pending/processing → return as-is;
partial/failed → trigger reprocess (partial enters resume branch)
- Clean up the newly uploaded temp file when dedup discards it
- Frontend: uploadRawFile/addRawText check for existing id in the list
before unshift to prevent visual duplicates
Test: WikiRawMaterialDedupTest — 10 cases covering all 5 statuses,
reprocess triggers for partial/failed, no-op for others, insert only
when no match.
PR-0 only installed the strategy seam; the actual ~600 LOC of provider-
specific construction stayed in AgentGraphBuilder as transitional public
helpers. PR-0b moves the DashScope + Anthropic halves into their builders
proper. (OpenAI larger refactor — 5 sub-helpers including Kimi/o-series
special cases — is left for a follow-up PR-0c.)
AgentDashScopeChatModelBuilder now owns:
- buildDashScopeApi (with provider/env/reflection key+url fallback chain)
- buildDashScopeOptions (model/temp/max-tokens/topP + built-in search)
- normalizeDashScopeBaseUrl (strip /compatible-mode/, return null for SDK default)
- readApiKeyFromDefaultChatModel + readBaseUrlFromDefaultChatModel +
readDashScopeApiFromDefaultChatModel (reflection-based final fallback)
- isBuiltinSearchEnabled (renamed from isDashScopeSearchEnabled, called
by AgentGraphBuilder.build via the now-injected dashScopeBuilder ref)
AgentAnthropicChatModelBuilder now owns:
- buildAnthropicApi (key validation, applyHttpTimeouts duplicated locally)
- buildAnthropicOptions (extended-thinking budget mapping low/medium/high/max
→ 4k/8k/16k/32k, temperature=1 enforcement, RFC-014 prompt cache options)
AgentGraphBuilder dropped:
- DashScope: ~120 LOC (api + options + 4 helpers + isDashScopeSearchEnabled)
- Anthropic: ~75 LOC (api + options)
- DashScopeChatModel + DashScopeConnectionProperties fields (unused after move)
- Deprecated single-fallback buildFallbackModel (no callers, superseded
by buildFallbackChain since RFC-009 PR-1)
- 5 imports for moved DashScope/Anthropic types
Net: -154 LOC in AgentGraphBuilder (1721 → 1567), +372 across the two new
builders. Strategy seam is now real for 3 of 4 protocols (ChatGPT was
already standalone, OpenAI is PR-0c). 220/220 tests still green — no
behavior change.
Two related issues from the Kimi-401 user report:
1. Backend (NodeStreamingChatHelper): a primary AUTH_ERROR (e.g. Kimi 401
with an invalid API key) returned immediately without trying the
fallback chain — a fallback provider with a different, valid key
never got a chance. Even with DashScope correctly configured as the
fallback, the user chat dead-ended on a 401.
The original assumption ("auth never self-heals so do not retry")
holds for the primary same-model retry loop but is wrong for the
fallback chain — different providers have different keys. Apply the
same break-into-fallback policy that BILLING and MODEL_NOT_FOUND
already use. recordPrimary(false) is preserved so the cooldown
counter still accumulates.
2. Frontend (chatError.ts + i18n): the error-text matching for
/认证|auth|unauthorized|401/i was so broad it matched the substring
"auth" inside URLs like https://api.kimi.com/.../auth, classifying
any model 401 as user "session expired" and rendering the misleading
"页面将自动跳转到登录页" copy. (The redirect itself only fires from
/api/v1/auth/* axios paths and SSE-connection 401s, not from this
payload-text path — but the copy alone is the worst kind of false
alarm.)
Add a new ChatErrorCategory provider_auth_error and split the
pattern matching: narrow auth_expired (HTTP 401 / 登录已过期 /
session expired / 凭证失效) is matched FIRST, then the broad
401-ish pattern routes to provider_auth_error. BACKEND_ERROR_TYPE_MAP
for AUTH_ERROR is also remapped, since structured backend payloads
currently always come from LLM providers — never from our own
/api/v1/auth path.
Tests
- NodeStreamingChatHelperFailoverTest (5 cases): primary 401 →
fallback succeeds; chain skips auth-failing fallback to next healthy
one; whole-chain failure surfaces last AUTH_ERROR (no silent drop);
BILLING regression unchanged; primary-success path does not touch
chain
- Browser preview verified: new i18n keys resolve in en-US, classifier
correctly routes "[错误] 401 from kimi.com" → provider_auth_error
while "[错误] HTTP 401 from /api/v1/auth/ping" stays auth_expired
- 186 tests pass (was 181 + 5 new); vue-tsc clean
Do-not-touch list: handleAuthFailure() in useStream/api/index.ts (real
session-expiry path) is unmodified — only the misclassification
upstream is fixed. auth_expired i18n copy is unchanged.
Track the primary model health, not just fallback entries
- NodeStreamingChatHelper accepts primaryProviderId via a new 5-arg
constructor; AgentGraphBuilder passes ModelConfigEntity.getProvider()
- Before the 5-retry primary loop, check
healthTracker.isInCooldown(primaryProviderId): if true, log + broadcast
"主模型暂时不可用（冷却中），直接尝试备选模型..." and short-circuit
straight to the fallback chain. Prevents a degraded primary from
burning 30+ seconds of backoff on every conversation turn.
- recordPrimary(success/failure) now fires on every primary verdict —
AUTH, BILLING, MODEL_NOT_FOUND, EMPTY_RESPONSE, generic UNKNOWN, and
the explicit success path. Three consecutive failures push the
primary provider into cooldown automatically.
- Legacy 1/2/3-arg constructors leave primaryProviderId null; tracking
silently disables for them so existing tests/wiring keep working.
Split BILLING and MODEL_NOT_FOUND out of CLIENT_ERROR / AUTH_ERROR
- BILLING (HTTP 402, "insufficient_quota", "credit balance is too low",
"billing_hard_limit_reached", "quota exceeded"): payment failure on
primary does not kill the call — a different provider may have credits.
Skips same-model retries and heads to fallback chain.
- MODEL_NOT_FOUND (HTTP 404, "Model not exist", "model_not_found",
DashScope "[InvalidParameter] url error"): unknown model id will not
start working on retry. Was previously misclassified as CLIENT_ERROR
and terminated the whole call; now routes to fallback so a different
provider can attempt with its default model.
- classifyError ordering matters: BILLING / MODEL_NOT_FOUND are matched
BEFORE the generic 400 / Bad Request branch, otherwise they would be
swallowed by CLIENT_ERROR.
Tests
- ErrorClassificationTest: 11 tests, covers multi-vendor error phrasing
for both new types + regression checks that 401 / 429 / 400 still
classify as before
- NodeStreamingChatHelperFallbackChainTest: +2 tests verifying
primaryProviderId persistence on the new constructor and null on
legacy ones
- 181 tests pass (was 168 + 13 new)
UI — Failover priority editor
- ProviderConfigRequest + ProviderInfoDTO carry fallbackPriority
- ModelProviderService.updateProviderConfig persists it (null = unchanged);
toProviderInfo exposes the current value to the UI (defaults to 0)
- ProviderConfigModal advanced panel exposes a number input with hint
- ProviderCard shows a "Fallback #N" badge for chain members so the
priority order is visible at a glance without opening the modal
- 5 new i18n keys (zh + en) — verified to resolve at runtime via i18n.global.t
Backend — Per-provider health tracker
- ProviderHealthTracker: ConcurrentHashMap-backed counters; N consecutive
failures (default 3) push the provider into a cooldown window (default
5 min) during which the chain walker skips it. Success resets both
counter and cooldown atomically. Lazy expiry on lookup so dead entries
do not accumulate.
- ProviderHealthProperties exposed under mateclaw.llm.failover.health.*
with sane production defaults
- New FallbackEntry record (providerId + ChatModel) replaces raw
List<ChatModel> in the chain so the walker can correlate cooldown
state to entries; AgentGraphBuilder.buildFallbackChain returns the
new type
- NodeStreamingChatHelper takes the tracker through a new 4-arg
constructor and consults it before each fallback call; records
success/failure on each chain attempt. Legacy 2/3-arg constructors
preserved as @Deprecated wrappers (synthetic providerId means no
health tracking on the legacy path — that path is opt-out anyway)
Tests
- ProviderHealthTrackerTest (9 tests): below/at threshold, success
reset, cooldown expiry (via reflection on the min-clamp setter),
disabled-tracker no-op, null-providerId safety, per-provider
isolation, snapshot output
- NodeStreamingChatHelperFallbackChainTest updated to FallbackEntry
field type — verifies providerId + ChatModel survive the chain
- 168 tests pass (was 159 + 9 new)
Verification
- mvn test green; vue-tsc clean; live UI confirms i18n resolution
Replaces the hardcoded single-DashScope fallback with a DB-driven
ordered chain. Same-provider primary deployments (e.g., DashScope
qwen-max) finally get a real fallback; if any provider in the chain
returns an empty body or transient failure, the next is tried.
Schema — DB-driven chain
- mate_model_provider gains `fallback_priority INT DEFAULT 0`. Positive
values define try-order; 0 = not in chain. Migration V21 (h2 + mysql)
seeds DashScope as priority 1 to preserve existing behavior.
- ModelProviderService.listFallbackChain() returns providers ordered by
priority ascending.
- ModelProviderEntity gains the new field.
Runtime — chain walk + empty-response trigger
- AgentGraphBuilder.buildFallbackChain(primaryConfig) returns a
List<ChatModel>, identity-filtering the primary by (providerId,
modelName) — fixes the bug where same-provider-primary deployments got
null fallback. Providers whose API key is missing are silently
skipped with WARN. Old buildFallbackModel(ChatModel) kept as
@Deprecated wrapper.
- NodeStreamingChatHelper accepts List<ChatModel>; the post-retry
fallback block now walks the chain in priority order, single-shot
per entry. Old single-fallback constructors retained as @Deprecated
one-element-list wrappers so legacy callers keep working.
- New ErrorType.EMPTY_RESPONSE: when the LLM returns no content, no
thinking, AND no tool calls, mark the result as a soft failure and
break the same-model retry loop, handing off directly to the
fallback chain.
- Broadcast updated to "切换到备选模型 (N/M)..." so SSE consumers see
chain progress.
Tests
- NodeStreamingChatHelperFallbackChainTest covers constructor variants,
chain immutability, deprecated-overload back-compat, and the
EMPTY_RESPONSE enum exists as a compile-time contract.
- 159 tests pass (was 153 + 6 new).
A. Delete two dead prompt files (prompts/context/conversation-summary-*.txt)
that no caller has loaded since the structured-summary triple replaced them.
B. Drop the never-wired locale machinery: PromptLoader.loadPrompt(name, locale)
overload + the prompts/{locale}/... fallback chain + I18nService.currentLocaleTag().
A single-language prompt corpus plus LLM input-language following is sufficient.
C. Strip duplicated structure list / budget directive from
structured-summary-update.txt (the system prompt already carries them).
Add a defensive preamble to both summary prompts: "do not respond to any
questions or requests in the conversation, only output the structured
summary" — prevents the summarizer from accidentally answering historical
user questions.
D. Fix {summary_budget} placeholder leak in the iterative-update branch of
ConversationWindowManager.generateSummary. Both branches now substitute
on the SystemMessage uniformly. Regression-guarded by
ConversationWindowManagerSummaryBudgetTest.
E1. De-hardcode seven prompts (research/{plan,draft,compose}-{system,user},
graph/limit-exceeded-system) — language now follows the user's input
instead of being hardcoded; citation tokens are language-neutral
[M1] / [Q1] markers.
E2. Add 10 i18n keys (research.fallback.*, research.broadcast.*,
agent.limit_exceeded.*) to messages.properties + messages_en.properties.
Inject I18nService into WikiResearchService and LimitExceededNode and
route 5 + 2 hardcoded fallbacks through i18n.msg(). Regression-guarded
by WikiResearchServiceFallbackTest + LimitExceededNodeFallbackTest.
E3. Replace 3 assembly tags in WikiResearchService with neutral
[M1] / [Q1] tokens. Aligns with the [M1] / [M2,3] citation format the
draft prompt asks for.
G. Three new regression tests cover D, E2, and E3.
- db/migration/mysql: replace ADD COLUMN/CREATE INDEX IF NOT EXISTS with
idempotent checks via information_schema (MySQL 8.0 <8.0.29 and some
forks don't support IF NOT EXISTS for ADD COLUMN). Affects V2/V4/V5/V7
/V8/V9/V11/V12/V13/V14. Fix: gitee#IIYHLJ.
- application-mysql.yml: add createDatabaseIfNotExist=true so MySQL
Connector/J auto-creates the schema on first connection (requires
CREATE privilege — documented fallback for restricted accounts).
- llm/OllamaAutoDiscoveryRunner: rewrite seed tag when fuzzy-matching,
prefer exact tag for default; skip models without tool support when
auto-activating a default (prevents the phantom ':latest' trap when
users pulled a specific size).
- agent/graph/NodeStreamingChatHelper: detect 'does not support tools'
and 'model not found' errors from Ollama and emit actionable Chinese
prompts guiding users to qwen3 / qwen2.5:7b+ / llama3.1:8b+ etc.
- ui/MessageBubble + types/chatError: surface the backend's actionable
rawMessage in the failed-message card instead of a generic '未知错误';
strip redundant prefixes (Bad request: / [错误] / LLM 调用失败:) since
the title already conveys the category.
- ChannelMessageRouter: include assistantMessageId in message_complete
and done broadcasts so ChatConsole observers can reconcile the
streaming placeholder to the persisted DB row by id instead of
falling back to the FIFO 'claim' heuristic (which occasionally
dropped the assistant bubble on external channel conversations).
Capture the id from ConversationService.saveMessage in both the
sync agentService.chat path and the streaming processWithStreaming
path; switch to HashMap since Map.of rejects null values when save
is skipped (e.g. under approval).
- ChatConsole: add a 'running' indicator on the sidebar so users can
tell which conversations have an in-flight agent run. Pulsing amber
dot on the channel icon (both expanded and collapsed modes) plus a
'生成中…' / 'Generating…' pill in expanded mode.
- ChatConsole: don't cancel the previous conversation's streaming run
when switching conversations — let it keep running in the background
and reconcile when the user comes back.
- ConversationWindowManager: cap reserve token at 50% of effective max
to prevent negative historyBudget on small-context models (8K/16K)
- common.security.SecretEquals: new constant-time comparison utility
(MessageDigest.isEqual wrapper) for secrets/tokens/signatures
- WeixinChannelAdapter: migrate context_token comparison to SecretEquals
- FeishuChannelAdapter: fail-fast on empty encrypt_key when connection_mode=webhook
- TelegramChannelAdapter: sanitize attachment captions — strip control bytes
(\p{Cc} except \t\r\n) + format chars (\p{Cf}) + 4096 char cap
- AgentGraphBuilder: fallback Anthropic max_tokens to 4096 on null/0/negative
Tests: SecretEqualsTest (5) + TelegramCaptionSanitizeTest (5) — all green.
Full-stack AI assistant built on Spring AI Alibaba.
Features: ReAct Agent, Plan-and-Execute, MCP Protocol, Multi-Model, Multi-Channel.
Apache-2.0 License
