The knowledge-base trust verification recorded wiki citations from every
non-readFile tool response by sniffing its JSON for a top-level title /
pages / chunks field. Tools like getGoalStatus return a top-level title,
which falsely populated the citation set and then forced [n] citations on
the final answer (otherwise flagged EVIDENCE_INSUFFICIENT). Gate citation
mining on the wiki_* tool name instead.
Likewise, the grounded answer contract (cite-or-refuse) was appended to
every ReasoningNode call unconditionally, degrading general agents that
have no knowledge base. Append it only when the agent has a wiki_* tool
bound, scoping the strict regime to KB-grounded scenarios.
Adds a regression test asserting a non-wiki tool with a top-level title
creates no wiki citations.
- rescan: keep the KB id as a string end to end so the 19-digit snowflake id
isn't truncated past Number.MAX_SAFE_INTEGER (rescan no longer 404s)
- lint: resolve [[...]] targets against page slugs AND titles like the viewer,
so a title reference to an existing page is no longer reported broken
- ingest: derive the slug deterministically from the title (no inconsistent
romanization), auto-recompute broken links once a KB finishes importing,
and reconcile dangling [[concept]] links — redirect to the covering page via
declared aliases, or demote to plain text when uncovered
- add the page aliases column migration for h2 / mysql / kingbase
Surface the connected database product as a subtle chip in the Dashboard
header. SystemHealthService now reports a database label on /system/health
(reused by the front-end — no extra request), derived from a new
DatabaseBootstrapRunner.getDatabaseLabel() that reads the JDBC product name
once and normalizes it to a canonical label (MySQL / MariaDB / PostgreSQL /
H2, and 人大金仓 for the KingbaseES family), collapsing driver version noise.
Add ErrorClassificationTest regression cases for the AI-gateway retry
hardening: 5xx-before-4xx ordering (a proxy 502 whose body says
"bad request" stays retryable), Chinese / numeric provider billing
patterns, and DNS / TLS infrastructure-fatal detection.
Fix the cert-trust pattern while adding its test: Java's ValidatorException
emits "PKIX path building failed" with an uppercase PKIX and the error
chain is not lower-cased, so the previous lowercase pattern never matched
— an untrusted/expired cert chain fell through to the retryable
SERVER_ERROR bucket and was retried in vain instead of failing over.
The disk fallback's catch block logged at debug, so a failed scan
silently dropped recovered files — reproducing the same 'bot can't see
the file' symptom the fallback was added to fix. Promote to warn with
the full stack trace, matching cacheRecentFile's logging.
openAiCompatibleClientBuilder() was creating a new java.net.http.HttpClient
per request. Each instance spawns a selector thread and connection pool that
are never closed, exhausting the OS thread limit under frequent model-test
calls (e.g. DeepSeek provider).
Elevate the HttpClient to a static singleton so all OpenAI-compatible
provider requests share one connection pool and one selector thread.
Closesmatevip/mateclaw#328
loadRecentFilesFromDisk now filters out files older than RECENT_FILE_TTL_MINUTES
(60 min) so the disk fallback matches the Caffeine cache TTL and does not inject
stale attachments into future conversations.
Testability refactoring:
- recentFileCache: private → package-private (tests can seed the cache directly)
- chatUploadsRoot: new package-private Path field (tests redirect to @TempDir)
- loadRecentFilesFromDisk: add (Path dir, long cutoffMs) package-private overload;
private (String) wrapper delegates to it
- injectRecentFiles: private → package-private
New test class FeishuRecentFileCacheTest (14 cases):
- loadRecentFilesFromDisk: non-existent dir, empty dir, fresh files sorted
newest-first, stale files excluded by TTL, mixed fresh+stale, >5 files capped,
timestamp-prefix stripping, MIME guessing from extension
- injectRecentFiles: Caffeine cache hit, cache-miss disk fallback, empty disk,
duplicate-path dedup, image vs file part typing, null textContent guard
Relates to #325
The per-chat recent file cache (Caffeine, 60 min TTL) is purely
in-memory. After a process restart, GC eviction, or TTL expiry the
cache is empty, but the staged copies under data/chat-uploads/ survive
on disk. A follow-up text message that should have seen the cached file
instead found nothing — the bot replied as if no file was ever sent.
Changes:
- injectRecentFiles(): fall back to scanning data/chat-uploads/{id}/
when the Caffeine cache misses, sorted by last-modified time, capped
at RECENT_FILE_MAX_PER_CHAT (5).
- cacheRecentFile(): promote catch log from debug → warn with full
stack trace so silent download failures are visible in production
logs. Add entry-level info log for correlation.
- New helper loadRecentFilesFromDisk() + guessContentType().
Closes#325
Relates to #201
The KingbaseES change removed usingDbTime() unconditionally, which made
every deployment (MySQL/H2/PostgreSQL) fall back to app-server time for
distributed lock timing — reintroducing node clock-drift risk in
multi-instance setups. Re-enable usingDbTime() for databases in ShedLock's
built-in dialect map and skip it only for KingbaseES, which is not covered
and would otherwise throw at lock acquisition.
Add KingbaseES support as an opt-in profile: dedicated migration tree, bilingual seed data, runtime DbType detection (KINGBASE_ES / POSTGRE_SQL), and JDBC URL handling in the datasource manager.
* feat(wiki): unify raw materials & source watcher into a Sources tab with per-KB auto-sync
The raw-material directory scan and the Advanced "source watcher" sub-tab were
the same engine (same kb.sourceDirectory, same WikiDirectoryScanService) split
across two surfaces with two editable directory inputs. Merge them into one
"Sources" tab (upload / paste / directory manual scan + auto-sync toggle +
the raw-material list) and drop the watcher sub-tab from Advanced.
Auto-sync is now per-KB opt-in: a new watcher_enabled column (V146) gates the
periodic scan per knowledge base. The server-global mate.wiki.watcher-enabled
stays as an ops master switch — a KB is auto-scanned only when both are on
(AND). Manual scans are unaffected. Scan interval stays global for now
(tracked separately).
Closesmatevip/mateclaw#314
* docs(wiki): document source-watcher global switch env vars
Expose MATE_WIKI_WATCHER_ENABLED / MATE_WIKI_WATCHER_INTERVAL_MS as
explicit placeholders in application-mysql.yml, .env.example and
docker-compose.yml, mirroring MATE_WIKI_ALLOWED_SOURCE_ROOTS. Notes the
AND semantics (global ops gate + per-KB toggle) so operators know the
global switch alone is not sufficient.
- Fail closed to a global fallback sandbox root when a conversation has no
per-workspace base path, instead of leaving file/shell tools unconstrained
- Refuse shell commands that delete the workspace root directory itself
- Block workspace-boundary escapes at the policy layer before the approval
prompt, not only at execution time
- Approval bar now shows the actual command / target path being approved
Closes#289 — after an MCP server (re)connects, chat queries kept replying
"from memory" instead of calling MCP tools.
Root cause: agents snapshot their tool set at build time and are cached in
AgentService.agentInstances, but MCP server lifecycle changes never
invalidated that cache (unlike model-config / tool-guard changes which do).
A stale, tool-less agent graph survived until process restart.
Changes:
- Add McpServerChangedEvent; McpServerService publishes it on connect /
disconnect / reconnect / delete / (re)connect-failure / batch refresh /
startup init. AgentService listens and calls refreshAllAgents(), so the
next turn rebuilds against the live MCP tool set. Also closes the boot
race where the web server accepts requests before the @Order(200) MCP
init runner finishes.
- Make create/update/toggle connect asynchronously on a dedicated pool
("mcp-connect") so a slow/unreachable server can no longer freeze the
admin request; status returns immediately as "connecting".
- UI: render the new "connecting" status (pulsing amber dot), show a
friendly "connecting in background" toast, and poll until the status
settles (window widened to ~40s to outlast the default connect timeout).
- UI: MCP config modal no longer closes on outside/backdrop click — only
the × and Cancel buttons close it, so an accidental click can't discard
unsaved config.
Verified E2E: ckjia-shopping (参考价) MCP server connected at runtime with
no backend restart; the cached 通用助手 agent immediately enabled and called
ckjia_shopping_recommend, returning real product cards.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The webchat conversationId (webchat:<key8>:<visitorId>[:<sessionId>]) and the
derived username (webchat:<visitorId>) are written to VARCHAR(64) columns, but
visitorId had no validation and sessionId allows 64 chars — so a long visitorId,
or a legitimate 64-char sessionId, overflows the column and the getOrCreateConversation
INSERT throws (500 on /stream). Validate visitorId (charset + blank->UUID) and
fold the variable part into a stable hash when the derived id/username would
exceed 64 chars, keeping short ids byte-identical (backward compatible). Also
make listSessions filter on exact owner username, not just the conversationId
prefix, so system-owned rows can never leak via a crafted visitorId. Adds
boundary regression tests.
Add per-visitor session management for the WebChat Web/API access mode:
list a visitor's conversation threads, fetch a thread's messages, and
delete a thread.
Authorization: visitorId is a client-asserted request param, so it cannot
be trusted on its own — deriving conversationId from it and then checking
ownership against it is tautological (any caller passes). Instead, /stream
issues a per-visitor token = HMAC-SHA256(jwtSecret, channelId:visitorId),
returned in the meta event; the management endpoints require it back via
the X-MC-Visitor-Token header and verify it in constant time. The signing
secret is server-only (unlike the public channel API key) and the channelId
in the payload makes tokens non-portable across channels.
Includes regression tests for token issuance/verification semantics
(forged visitorId rejected, cross-visitor and cross-channel tokens rejected,
tampered tokens rejected).
The WebChat stream endpoint resolves agentId only when the
(visitorId + sessionId) conversation is first created; later requests
with a different agentId reuse the existing conversation's agent and
silently ignore the new value. Document this on WebChatRequest.agentId
so integrators know to use a new sessionId to reach another agent.
The vision/video sidecar dropdown filtered candidates through the built-in
capability heuristics, so provider-compatible models whose custom names match
no known prefix (and carry no declared modalities) were hidden and could not be
selected as a sidecar — even when they natively support the modality.
- listByType now returns every enabled chat model, annotating each row with a
transient modalityCapable flag and sorting known-capable rows first, instead
of hard-filtering recognized models only.
- MultimodalRouter honours an explicitly configured sidecar model instead of
dropping it when the heuristics don't recognize it; a wrong pick degrades
gracefully through the caption path rather than silently disabling routing.
- ModelPicker gains an optional capability badge; the sidecar UI tags
recognized vision models while keeping every enabled model selectable.
The mention alias-learning fed every identifier of every mention in a
delivery into the per-chat alias cache. A single delivery of "@bot @alice"
matched the bot by its global id and then learned alice's openId as a bot
alias, so every later "@alice" message was misdetected as @bot and the agent
replied to messages never addressed to it.
Only single-mention deliveries are unambiguous bot identities, so restrict
alias learning to them — a multi-mention delivery mixes the bot with
co-mentioned humans, and Feishu's dual-delivery alias form is itself a single
mention, so this is safe and keeps the learning feature working. Also cap the
per-chat alias set size. Adds a [bot, human] co-mention regression test.
Rename the webchat channel from "embed widget" to "Web / API access" (key
unchanged, docs/i18n only) and extend the backend SSE endpoint for pure
backend integration:
- WebChatRequest gains optional agentId (route one Key to multiple agents;
rejected unless the agent shares the channel's workspace) and sessionId
(one visitor, multiple isolated threads).
- sessionId is validated ([A-Za-z0-9_-]{1,64}) and only composed into the
server-derived conversationId — raw conversationIds are never accepted, so
the key+visitor namespace still bounds every thread.
- A `meta` SSE event echoes the effective sessionId/conversationId at stream
start so callers can persist and re-address a thread.
- Memory stays attributed per visitor (api:<visitorId>), shared across that
visitor's sessions.
All new fields are optional; omitting them reproduces the prior behaviour
byte-for-byte. Refs matevip/mateclaw#295.
ChannelMessageEventBridge.onChannelMessage() ran synchronously on the Lark
SDK WebSocket dispatch thread; when the DB pool was saturated its ingest
query blocked that thread and subsequent messages were silently dropped.
- @Async moves trigger ingest onto the (vthread + SecurityContext-propagating)
async executor, freeing the WS dispatch thread.
- Add a 10s timeout to the refreshTenantAccessToken() HTTP request, which
previously had none.
Scope narrowed per review to only fix#208: the @mention alias learning and
session-id changes are dropped (to be raised as separate PRs), and dev's
existing message dedup is left untouched.
Closes#208
Add an execute_code built-in tool that runs python/bash/node code the agent
writes on the fly, so a documentation-only skill (a SKILL.md with no bundled
scripts) can be acted on. Scoped runs inject the skill's secrets and run in
the skill directory; otherwise a private scratch directory is used. Host
secret env vars are scrubbed from the subprocess. execute_code is an
agent-wide capability, registered in the tool catalog (V143), and screened
by the tool guard with a dedicated set of destructive-pattern rules.
Tests cover python/bash/node execution, scratch-dir fallback, env scrubbing,
argument decoding, and guard gating.
- WikiPageTypeProfile: normalise pageType keys to lowercase on set, so a
user-authored profile with an uppercase key still matches the
case-insensitive hasPageType/get lookups.
- WikiDirectoryScanService: normalise the symlink-resolved glob base to
forward slashes so directory-scan globs work on Windows paths.
- Agents roster tag filter: keep selected tags that no longer exist on any
agent visible and deselectable (and show the filter bar when only such
orphan selections remain) instead of silently filtering with no way to clear.
A second POST to /reclassify on the same KB spawned an independent pass over
the same pages, doubling LLM spend and racing the first pass's page-type
writes. Add a per-KB in-flight guard that rejects a concurrent run with a
friendly message (409 via R.fail rather than a generic 500), released in a
finally once the async pass completes. Also count and broadcast per-page
failures so an all-failing run is visible instead of reporting changed=0, and
type the api modelId param as string|number per the snowflake ID convention.
The resilient connect() override started the child process into a local
variable and never set the parent's private process field, so the parent's
closeGracefully() logged "Process not started" and left the child running on
every reconnect/disable. Retain the started process and override
closeGracefully() to destroy it and interrupt the spawned I/O threads.
Also make the inboundSink/errorSink reflection fail loud (the transport is
useless without them — a silent null yielded a connected transport that timed
out on every call), force UTF-8 on the stdio reader/writer, honour a
cooperative closing flag in the reader loops, and drop duplicate PATH entries.
Add a backfill path so pages created before a KB's pageType profile changed
can be migrated into newly-added types. A per-page classify-only LLM call
(title + summary in, single page_type out) is normalised through the profile
and written back via a partial update that never touches page content.
Exposed as POST /knowledge-bases/{id}/reclassify (admin) and a "re-classify
existing pages" action in the Wiki advanced panel.
Wiki page classification was only profile-aware in the main ingest pipeline.
Transformation outputs hard-coded "synthesis", agent-created pages were left
untyped, and the frontend hard-coded the built-in ten types for ordering,
colouring and labels — so custom/synthesis types sank to the bottom, rendered
grey and showed raw keys.
Backend:
- Add nullable target_page_type column to mate_wiki_transformation (V142,
mysql + h2) plus the entity field and CRUD normalization (blank = use
profile fallbackType; membership validated at save time, not edit time).
- Route transformation single-run + KB-aggregate page saves through
WikiPageTypeProfileService.normalizePageType so output joins the KB
classification; agent wiki_create_page now lands on the profile fallbackType
instead of an untyped page.
Frontend:
- Load + parse the KB pageType profile into the wiki store (order, labels,
fallbackType) on KB select / refresh.
- New useWikiPageType composable: profile-driven label (3-tier fallback) and
colour (built-in fixed + deterministic hash palette for custom types).
- Sidebar grouping order, graph colouring, node panel, graph filter and the
page header badge now follow the profile; transformation editor gains a
target-type dropdown sourced from the profile when output target is a page.
Refs #292
Add 5-arg buildContextMessage overload that emits [system-context] Model:
for every origin (web/cron/IM). Legacy 3/4-arg overloads delegate to the
new one with null model args, keeping their output byte-identical.
Also fix pre-existing FeishuMentionTest compile error caused by removed
mentionMatchesAnyAlias/collectMentionIdentifiers methods.
A silent (no-op) cron run still makes a full LLM call, but the marker
message was persisted via the token-free saveMessage overload, so the
settings-page Token total (which aggregates MessageEntity token columns)
under-counted cron spend by exactly the silent runs.
Carry chatResult's prompt/completion tokens onto the marker message,
matching the non-silent branch. Closes#284.
The base directory is canonicalized via toRealPath before walking, so the
walked files carry the symlink-resolved prefix. The PathMatcher was built
from the literal pattern, so a symlinked base never matched and files were
silently dropped. Rebuild the glob against the resolved scan root, escaping
glob metacharacters in the base so a real directory name containing */?/{}/[]
is treated literally.
The upstream StdioClientTransport breaks out of its inbound read loop on
any JSON parse error, permanently killing the reader thread. Some MCP
servers write non-JSON debug output to stdout (e.g.
"=== Document parser messages ==="), which triggers this and causes all
subsequent valid JSON-RPC responses to be lost → 30s timeout → agent stuck.
Override connect() in CwdAwareStdioClientTransport to skip non-JSON lines
(log at DEBUG) instead of breaking, keeping the inbound thread alive.
Closes#226
The #254 fix changed resolveConventionPath to {name}-{hashCode} and folded
hyphens to underscores, which re-pathed every existing skill (browser-cdp ->
browser_cdp-<hash>) with no migration, orphaning already-created workspaces and
breaking SkillWorkspaceManagerApplyBundleTest. Drop the hash suffix and keep
hyphens: the bare Unicode-preserving sanitized name already prevents the
non-ASCII collision (distinct CJK names map to distinct dirs) and leaves ASCII
kebab-case paths identical to the legacy scheme. Add path regression tests.
Skills with non-ASCII (e.g. Chinese) names collapsed to underscores in the workspace path, so two such skills resolved to the same directory and overwrote each other. Preserve Unicode letters/digits when sanitizing the path so distinct names map to distinct directories. Also write SKILL.md with CREATE_NEW to avoid a TOCTOU race on concurrent uploads, and mount the skills workspace as a named Docker volume so packages survive container restarts.
Closes#254
updateTextContentFromScan / updateBinaryFileFromScan are only reached via
self-invocation from the ingest* methods, so the proxy-based @Transactional
never applied. Each runs a single atomic updateById; remove the misleading
annotation and document why.
Directory-scan ingestion deduped only by content hash, so when a file at a known path changed, the new hash missed the existing raw and a second row was inserted for the same source_path — both rows then generated wiki pages, accumulating duplicates. Make source path the primary dedup key: same path + same hash skips, same path + changed hash updates the existing raw in place (reset to pending, re-process), falling back to the content-hash check only for genuine copies at new paths. Reprocessing reuses the same rawId, so deleteExclusiveBySourceRawId cleans the old pages before regeneration — no duplicate rows and no duplicate pages. findBySourcePath gains LIMIT 1 to tolerate pre-existing duplicates; docs/fix-duplicate-raws.sql remediates existing data.
Closes#271
Gate MarkdownNormalizer behind mate.agent.markdown-normalize-enabled (default
true) so operators can disable the rewrite verbatim if a normalization edge
case ever mangles a legitimate answer.
LLMs routinely emit malformed Markdown (missing heading spaces, glued `---`, unaligned table pipes) that prompt rules cannot reliably prevent. Add a zero-token, regex-only MarkdownNormalizer applied on the FinalAnswerNode convergence path before persistence / channel delivery. It is code-fence aware, idempotent, and conservative (em-dash `---`, `#5`-style refs, stray prose pipes are left untouched). RETURN_DIRECT verbatim output and approval-wait paths return earlier and are unaffected.
Closes#274
Same-message attachment reads failed on Feishu: the content part carried the sandbox-external ~/.mateclaw/media/ path, so read_file/extract_document_text were rejected by WorkspacePathGuard and ChatUploadResolver's basename fallback missed (media names {messageId}_{key} vs chat-uploads {millis}_{fileName}). cacheRecentFile already copies the attachment into the per-conversation data/chat-uploads/ dir; return that absolute path and stamp it onto the current message's image/file/audio/media parts so the path surfaced to the LLM is resolver-reachable. Reuses the existing copy — no extra I/O, no new state. Rich-text post multi-image and other channels sharing the media dir are noted as follow-ups.
Closes#278
Allow knowledge base source paths to be configured as a newline-separated
list of absolute paths or glob patterns rather than a single directory.
- Each non-blank, non-# line is treated as one path or glob pattern
- Plain paths (no wildcards) retain the existing recursive-scan behaviour
filtered by SUPPORTED_EXTENSIONS
- Glob patterns (e.g. /data/ocr/**/*.txt) walk from the fixed-prefix base
and apply Java's PathMatcher against each candidate's absolute path
- Patterns whose filename segment explicitly specifies an extension
(*.txt, *.{xlsx,csv}) skip the SUPPORTED_EXTENSIONS secondary filter,
respecting the user's explicit choice (key for OCR pipelines that
produce .txt output and should ignore the original PDF scans)
- Candidates collected across multiple patterns are deduplicated by
resolved absolute path so overlapping patterns don't double-count
- Symlink-escape check uses each pattern's own validated scan root
- parseSourcePatterns / extractBasePath moved into WikiSourcePathValidator
to eliminate a static circular reference between the two services
- Frontend watcher panel: single-line <input> replaced with <textarea>
supporting multiline editing; i18n updated with example patterns
- No DB schema change; fully backward-compatible with existing single-path
configs stored in sourceDirectory
Closes #(pending issue)
finishRunAndPublish updated the run row with status and finished_at but dropped the LLM token usage, so mate_cron_job_run.token_usage stayed NULL and the scheduler history always showed 0. Write the prompt+completion token total (already computed for the assistant message row) to the run row, guarded so non-LLM paths (reminder direct-push with a null ChatResult) leave the column untouched.
Closes#262
The skill meta-tools (listAvailableSkills, readSkillFile, listSkillFiles, load_skill, runSkillScript) queried the full skill catalog without checking the calling agent's bindings, so an agent scoped to a subset of skills could still read, load, or execute any skill via direct tool calls. Resolve the agent's bound skill ids from the tool context and filter/deny accordingly: a null binding set means no restriction (backward compatible), a non-null set (including empty) restricts access to that set — matching the system-prompt catalog filtering.
Closes#264
Surface directory-scan failures to the user via toast and render ScanResult.errors[]; expose MATE_WIKI_ALLOWED_SOURCE_ROOTS as a Docker env entry with blank-entry filtering in the path validator; set C.UTF-8 locale in the runtime image so non-ASCII file names decode correctly during scans.
Fixes#259
Add a single global-proxy switch that routes the backend's outbound traffic
through a configured HTTP/HTTPS/SOCKS proxy, for deployments that cannot reach
overseas APIs directly or must use a unified egress.
- ProxyManager installs the proxy via a default ProxySelector (honored by
java.net.http and HttpURLConnection), the proxy system properties, and a
--proxy-server arg for the browser tool; restores direct egress when
disabled. One switch covers LLM, web search, media generation, channels,
MCP and the browser.
- SOCKS applies to the HttpURLConnection-based egress only; the java.net.http
LLM/streaming path uses an HTTP proxy, and the UI states this.
- New Settings -> Network Proxy page: enable toggle, address, bypass list,
test-connection, and a coverage summary. Config stored as key/value in
mate_system_setting (no migration).
refs #109
Streaming chat ran render tools on an async thread with no bound request,
so download links lost their host and arrived without a domain. Resolve the
host on the request thread and carry it through ChatOrigin/ToolContext;
falls back to a configurable public-base-url, then a relative path.
Agents now have a per-agent primary wiki KB stored on
mate_agent.primary_kb_id. KBs remain workspace-shared — selecting one in
the agent editor only chooses the default wiki target for that agent, it
does not change the KB's ownership or visibility.
Backend
- AgentEntity: add primary_kb_id field (FieldStrategy.ALWAYS so the UI
can clear it back to "no primary")
- AgentController#update: switch body to Map<String, Object> so we can
tell "field missing" apart from "explicit null" via containsKey, then
convertValue back to AgentEntity
- WikiKnowledgeBaseService:
- new resolvePrimaryKb(agentId): prefers agent.primary_kb_id when it
points to a workspace-visible KB; falls back to legacy
kb.agent_id marker, then to most-recently-updated workspace KB
- listByAgentId now returns the full workspace set (KBs are
workspace-shared under the new model)
- update(id, name, description) no longer touches agent_id
- WikiController: new GET /knowledge-bases/bindable for the UI picker;
PUT /knowledge-bases/{id} no longer reads agentId
- WikiKnowledgeBaseEntity: add FieldStrategy.ALWAYS on embeddingModelId
and configContent so explicit nulls actually unbind/clear instead of
being silently skipped by MyBatis-Plus's NOT_NULL default
- Migrations V129 (H2 + MySQL): add primary_kb_id column + index, backfill
from legacy kb.agent_id, MySQL uses INFORMATION_SCHEMA guard +
PREPARE/EXECUTE for idempotency
- WikiKnowledgeBaseServiceTest: 13 cases, all passing
Frontend
- Agents.vue: new "Knowledge Base" tab, radio-select bindable KBs
- API: listBindableKBs() + Agent.primaryKbId typed string | number | null
- IDs handled as strings throughout (Snowflake-safe)
- i18n keys for the new tab in zh-CN and en-US
The previous private-repo support inlined the access token into the
clone URL and then logged that URL on success — leaking the token to
log files, container stdout, and any IOException thrown when the clone
failed. The token also appeared in the process command line, visible
to anyone with shell access via `ps`.
Switch to git's GIT_CONFIG_COUNT/KEY/VALUE environment variables, which
inject `http.extraHeader: Authorization: Bearer <token>` into the child
process without ever touching argv or the repo URL. The URL stays
pristine, so the existing INFO log and error message are safe.
Other changes:
- Resolve token from `mateclaw.skill.github-token` property first, then
fall back to GITHUB_TOKEN env var. Keeps the original deployment
contract while letting admins manage the credential via configuration.
- Tighten the host check (prefix match on `https://github.com/` etc.)
so a crafted URL like `https://evil.com/?u=github.com/...` cannot
trick the fetcher into forwarding the token to a third party.
- Set GIT_TERMINAL_PROMPT=0 so a bad token fails fast instead of
blocking on an interactive password prompt.
