MaybeCompact relied on RAM count which resets to 0 on restart (LoadFromDB
is a no-op). Messages accumulated in DB but never triggered compaction.
Now falls back to CountByKey DB query when RAM count is below threshold.
Replace channel column with context usage bar showing estimated tokens
vs compaction threshold. Token count estimated from messages JSONB
(octet_length/4 + 12K system prompt) — no migration needed.
Color-coded: green <60%, amber 60-85%, red >85%. Inline compaction
count with refresh icon.
* fix(channels): annotate DM messages with sender identity
Telegram and Zalo group messages already include [From: sender] prefix
so the agent knows who is talking, but DM messages were sent without
any sender context — the agent had no way to address the user by name.
- Telegram DM: add [From: @username] (or FirstName if no username)
- Zalo DM: add [From: displayName] when dName is present in payload
* fix(tests): add missing EnsureUserProfile to test stubs
AgentStore interface gained EnsureUserProfile in 4fce731 but the test
stub implementations were not updated, breaking CI on main.
---------
Co-authored-by: Luvu182 <208665161+Luvu182@users.noreply.github.com>
Replace hardcoded English tooltip strings in compaction config with
t() calls. Add insertContact, searchContacts, noContactsFound i18n
keys across en/vi/zh.
Add inline contact search above textarea in file editor (predefined
agents only). Search contacts and click to insert formatted snippet
at cursor position. Also make USER.md clickable in sidebar.
Extract KG from Memory page into standalone /knowledge-graph route.
Fix graph zoom: fitView after D3 force simulation settles so all nodes
are visible. Compact layout: merge stats/search/actions into single row,
inline filters with page header.
Add searchable contact dropdown in instances sidebar to find channel
contacts and add them as agent instances. Backend EnsureUserProfile
creates user_agent_profiles row on demand when admin adds contacts.
Regroup sidebar from 4 to 7 sections (Core, Conversations, Connectivity,
Capabilities, Data, Monitoring, System) for better organization. Keep
technical terms Nodes and TTS as-is across all languages.
Pass contactCollector through channel manager to all channel handlers
(Telegram, Discord, Feishu, Slack, Zalo) so contacts are automatically
collected when users interact with the agent.
The `allowEmpty` prop on `ProviderModelSelect` was defined but the
Select dropdown still auto-selected the first provider when the value
was empty, and there was no way to reset back to empty (auto-detect).
- Map empty provider to a sentinel so Radix Select can represent it
- Show an "(auto)" option in the dropdown when `allowEmpty` is set
- Pass `allowEmpty` from both embedding config callers
- Skip clearing the model on provider change in allowEmpty mode to
avoid a stale-closure bug where the second setState overwrites the
provider update
Co-authored-by: Luvu182 <208665161+Luvu182@users.noreply.github.com>
* feat(cron): configurable default timezone for cron expressions
Cron expressions (e.g. "0 8 * * *") are evaluated relative to a timezone.
Without an explicit per-job timezone, they default to the server's system
timezone, which may not match the user's local time — especially in Docker
containers (default UTC) or multi-region deployments.
This adds a `default_timezone` setting to `CronConfig` (IANA format, e.g.
"Asia/Ho_Chi_Minh") that is applied as fallback when a cron job has no
explicit `schedule.tz`. The setting is configurable via the UI config page
(Integrations → Cron Scheduler) and hot-reloads on config changes.
Backend:
- Add `DefaultTimezone` field to `CronConfig`
- Add `SetDefaultTimezone()` to `CronStore` interface + PG implementation
- Apply default TZ in `AddJob()` when `schedule.TZ` is empty
- Wire at startup + subscribe to config change events for hot reload
- Update cron tool description so LLM knows about gateway default
Frontend:
- Add timezone dropdown (20 common IANA timezones) to Cron config section
- Add i18n keys for en, vi, zh
* fix(cron): apply default timezone to existing jobs via computeNextRun
Pass defaultTZ as fallback to computeNextRun so existing cron jobs
(with timezone = NULL in DB) also use the gateway's configured default
timezone when computing next_run_at. This ensures old jobs benefit
from the timezone setting without needing a DB migration or backfill.
---------
Co-authored-by: Luvu182 <208665161+Luvu182@users.noreply.github.com>
* feat(providers): add Ollama local and Ollama Cloud provider support
Adds two new provider variants:
ollama — local/self-hosted Ollama instance
- Gated on providers.ollama.host in config (or GOCLAW_OLLAMA_HOST env)
- No API key required; Ollama's OpenAI-compat endpoint accepts any Bearer value
- Defaults to http://localhost:11434/v1 (configurable for LAN/remote hosts)
- Default model: llama3.3
ollama-cloud — Ollama Cloud (managed remote inference)
- Gated on providers.ollama_cloud.api_key (or GOCLAW_OLLAMA_CLOUD_API_KEY env)
- Bearer token from ollama.com/settings/keys
- Default base URL: https://ollama.com/v1 (overridable via api_base)
- Default model: llama3.3
Both variants use the existing NewOpenAIProvider (OpenAI-compat) — no new
provider struct needed. Both are registered from config file and DB (via
llm_providers table with ProviderOllama / ProviderOllamaCloud types).
OllamaCloud.APIKey follows all existing secret handling patterns:
MaskedCopy, StripSecrets, StripMaskedSecrets.
* feat(providers): wire Ollama into web UI and fix DB registration
- Add ollama + ollama_cloud to PROVIDER_TYPES constants (dropdowns)
- Fix setup wizard: skip API key requirement for Ollama local (isOllama)
- Fix bootstrap status: recognize Ollama local as no-API-key provider
- Add ollama_cloud to config-page KNOWN_PROVIDERS list
- Fix gateway_providers.go: move ProviderOllama before APIKey=='' guard
so DB-registered local Ollama providers actually register at startup
(same pattern as ClaudeCLI, which also needs no API key)
The onboard wizard sets GOCLAW_PROVIDER and GOCLAW_MODEL in .env for
initial bootstrap. Previously these env vars always overrode the config
file value via envStr(), making it impossible to change the default
provider/model through the Dashboard — every save was silently reverted
by ApplyEnvOverrides().
Change envStr to envFallback for these two fields: the env var is only
applied when the config file has no value (empty string). Once the user
saves a provider/model via the Dashboard, the config-file value wins.
Also:
- Stabilize ProviderModelSelect auto-select effect (useRef + useMemo)
- Add toast feedback on config save success/failure
Co-authored-by: Luvu182 <208665161+Luvu182@users.noreply.github.com>
Flush threshold (176K) was higher than compaction threshold (150K), so
flush was always skipped. Removed independent threshold calculation —
flush now always runs before compaction, gated only by enabled check and
per-cycle dedup guard. Also localized hardcoded compaction tooltips
(en/vi/zh), fixed inaccurate Memory Flush description, and removed dead
reserveTokensFloor/softThreshold UI inputs.
- Defer Telegram media download until after mention gate — pending
history now uses lightweight tags (no download) saving bandwidth
and avoiding errors on large files
- Fix compaction status query: has_summary now false when new messages
arrive after last compaction, re-enabling the compact button
- Fix HTTP compact endpoint: detach context from request so LLM
summarization isn't cancelled when browser closes connection
Backend defaults memoryFlush.enabled to true when config is nil
(memoryflush.go:39-46), but UI used `?? false` showing OFF.
Also fix sub-fields visibility to show when flush is defaulted ON.
Co-authored-by: Luvu182 <208665161+Luvu182@users.noreply.github.com>
Add Provider, Model, MaxTokens to PendingCompactionConfig so users can
override the LLM used for pending message summarization via the config
UI. Falls back to agent's provider/model when not set. Increase default
max_tokens from 512 to 4096. Add allowEmpty prop to ProviderModelSelect
to prevent auto-selecting first provider when empty means "use default".
When an announce (child trace) run starts, toggle the parent trace
status to "running" so the UI doesn't show "completed" with a running
child span. Restore to "completed" when the announce finishes.
Spans are now emitted BEFORE execution starts with status "running" and
input visible, then updated when the step completes. This lets users see
what the system is doing during long-running LLM calls and tool executions.
- Add EmitSpanUpdate() to collector with separate update buffer
- Flush ordering: batch INSERT new spans first, then process updates
- Split LLM/tool/agent spans into start/end pairs (agent loop + subagent)
- Emit tool span end inside goroutines for parallel calls (no orphans)
- EmitSpanUpdate is a channel send — works after ctx cancellation
- Fix compact endpoint using random provider instead of agent's configured provider+model
- Wire auto-compaction for all 5 channel types (telegram, discord, slack, feishu, zalo_personal)
via PendingCompactable interface and InstanceLoader
- Add global PendingCompactionConfig (threshold, keep_recent) to ChannelsConfig
- Wire global config through InstanceLoader and PendingMessagesHandler
- Increase compaction timeout from 45s to 180s for slow providers
- Add pending compaction config card to Behavior tab in config page
- Add HowItWorksCard (expanded by default) and toast notifications to pending messages page
- Add i18n support for all new strings (en/vi/zh)
- Add reviewer role to team system (backend + UI) for evaluate_loop workflows
- Fix handoff/evaluate_loop/delegate_search system prompt entries (were showing as custom tool)
- Filter inactive agents from delegation queries (DelegateTargets, SearchDelegateTargets)
- Fix agent link direction display (flip outbound→inbound when viewed from target side)
- Improve builtin tool seed descriptions with detailed action-oriented text
- Add i18n support for builtin tool descriptions (en/vi/zh frontend locale files)
- Notify LLM when KG extraction triggered on memory write
* fix(gateway): extract media forwarding helper and fix group writer cron/media gaps
- Extract appendMediaToOutbound() helper to deduplicate media attachment conversion across all outbound paths
- Add media forwarding to cron job handler, handoff announce, and teammate message flows
- Fix silent reply suppression to allow media-only responses (no text content)
- Skip group writer refusal prompt for system-initiated runs (cron, delegate, subagent) with empty senderID
* fix(agent): protect identity files in system-initiated group writer runs
- Move writer list lookup before senderID check so cron/delegate/subagent
runs still load group writer config
- Inject file protection prompt for system-initiated runs instead of
skipping entirely, preventing modification of SOUL.md, IDENTITY.md,
AGENTS.md, and USER.md
- Retain fail-open behavior when writer list is unavailable
---------
Co-authored-by: Nam Nguyen Ngoc <namnn.0911@gmail.com>
* fix(i18n): add missing config.json locale files
The config namespace was registered in i18n/index.ts but the actual
JSON files were never created, breaking the UI Docker build (tsc error).
Also fix .gitignore: config.json → /config.json so only the root
gateway config is ignored, not the i18n locale files.
Closes#107
* feat(ui): replace language cycle button with dropdown select
Users can now pick a language directly from a dropdown instead of
cycling through all options on each click.
* fix(i18n): keep "Subagent" in Vietnamese translations
"Agent con" and "sinh sản" sound unnatural; keep the English term.
* chore: align .dockerignore config.json pattern with .gitignore
Extract CompactGroup() from runCompaction() for shared use by both
auto-compact and HTTP compact endpoint. Wire provider registry into
PendingMessagesHandler so /v1/pending-messages/compact performs actual
LLM summarization (keep 15 recent + summary) instead of hard delete.
Add rows-affected guard in Compact() to prevent duplicate summaries
when concurrent compactions race on the same key.
Move count assignment before RAM trim in Record() so MaybeCompact
receives pre-trim count (51) that exceeds threshold (50), allowing
compaction to fire. Previously count was always ≤ limit after trim.
- Update go.mod and Dockerfile to Go 1.26
- Apply `go fix ./...` stdlib modernizations across 170+ files
- Add `go fix` to post-implementation checklist in CLAUDE.md
- Fix go fix misapplied rewrite in loop_history.go
- Add tool status display on channels during tool execution (streaming preview + reactions)
- Emit agent.activity events at phase transitions (thinking, tool_exec, compacting)
- Enrich delegation progress with per-member activity and tool info
- Add LLM-based intent classifier for DM status queries when agent is busy
- Keyword fast-path for cancel/status patterns (no LLM cost)
- Falls back to LLM classification with 5s timeout
- Supports status_query (immediate reply) and cancel (abort run) intents
- Register/unregister runs in makeSchedulerRunFunc for channel inbound tracking
- Add sessionRuns secondary index in Router for O(1) IsSessionBusy lookups
- Add intent_classify config toggle (global default + per-agent override)
- Add tool_status config toggle for channel tool status display
- Add i18n keys and translations (en/vi/zh) for status messages
- Add web UI config toggles for intent_classify and tool_status
* feat: add Z.ai provider support (general API + coding plan)
Add Z.ai (GLM) as a new LLM provider with two variants:
- `zai`: general API (api.z.ai/api/paas/v4)
- `zai_coding`: coding plan (api.z.ai/api/coding/paas/v4)
Reuses OpenAIProvider — Z.ai API is OpenAI-compatible with Bearer
token auth, SSE streaming, and reasoning_content support.
Includes: store constants, config struct fields, env var loading
(GOCLAW_ZAI_API_KEY, GOCLAW_ZAI_CODING_API_KEY), secret masking,
config + DB registration, onboard wizard, and UI provider types.
Default model: glm-5
Closes#100
* docs: add Z.ai provider entries to providers documentation
Remove the 700+ line interactive wizard (provider, model, channels, TTS,
features) since the web UI setup wizard already handles all of this.
The onboard command now only: prompts for Postgres DSN, tests connection,
generates gateway token + encryption key, runs migrations, seeds placeholder
providers, and saves .env.local with the 3 required secrets.
Delete 8 helper files: onboard_auto, onboard_verify, onboard_models,
onboard_resolve, onboard_claude_cli, onboard_tts, onboard_feishu, onboard_zalo.
- Add "do silently" instruction to BOOTSTRAP.md templates so agents
don't narrate internal file processing steps to users
- Rename channel instance form label from "Name" to "Key" for clarity
HandleMessage() always checked IsAllowed(senderID), which re-gates
individual senders against the allowlist even after the channel-specific
group policy (checkGroupPolicy) already approved access.
This caused group members who weren't individually listed in AllowFrom
to be silently blocked — even when GroupPolicy was set to "open" or the
group ID was in the allowlist.
Affected channels: Zalo Personal, Slack (any channel using HandleMessage
with peerKind "group"). Telegram was unaffected because it publishes
directly to the bus.
Now HandleMessage only enforces the allowlist for DMs. Group access
control is fully delegated to each channel's checkGroupPolicy.
Co-authored-by: Luvu182 <208665161+Luvu182@users.noreply.github.com>
- Move SOUL.md + IDENTITY.md to primacy zone (top of prompt) with
recency reinforcement at bottom to prevent persona drift in long chats
- Separate MCP tools into dedicated section with real descriptions
(truncated to 200 chars) instead of generic "(custom tool)"
- Add MCP tool priority instruction: prefer MCP over core tools
- Raise MCP inline threshold from 30 to 40 tools
- Add knowledge_graph_search to Memory Recall section (conditional)
- Add recency reminders for memory_search and knowledge_graph_search
- Add knowledge_graph_search to coreToolSummaries
- Recognize claude_cli and chatgpt_oauth as valid providers without API keys
- Skip API key validation and show CLI-specific UI when provider is claude_cli
- Clear API key state when switching provider type
- Update bootstrap status check to handle keyless provider types
Co-authored-by: Nam Nguyen Ngoc <namnn.0911@gmail.com>
- Add query="*" to list all entities for discovery
- Show top 10 entities as hints when search returns no results
- Display entity names instead of UUIDs in relation output
- Enable KG tool by default (admin still configures extraction settings)
- Tune force-directed graph layout: add forceX/forceY gravity to keep clusters compact
Fixes#94. The stream-json scanner never checked scanner.Err() after the
scan loop, so if a line exceeded the 1MB buffer limit the scanner would
stop with bufio.ErrTooLong and the response would be silently truncated.
- Check scanner.Err() after the loop and return an explicit error
- Increase max buffer from 1MB to 10MB to handle large tool outputs
Co-authored-by: Luvu182 <208665161+Luvu182@users.noreply.github.com>
* fix(mcp-bridge): add per-session agent context and HMAC verification
- Add per-session MCP config with X-Agent-ID/X-User-ID headers instead
of shared global config file
- Sign bridge context headers with HMAC-SHA256 to prevent forgery
- Add bridgeContextMiddleware to verify signatures on MCP bridge requests
- Store MCP configs in ~/.goclaw/mcp-configs/ outside agent workDir
- Use atomic writes (tmp + rename) for MCP config files
- Fix provider rename leaving ghost registry entries
- Remove provider_type from mutable fields on update
- Tighten temp dir permissions from 0755 to 0700
* feat(mcp-bridge): propagate channel routing context through MCP bridge
- Pass channel, chat_id, and peer_kind from agent loop to CLI provider options
- Inject X-Channel, X-Chat-ID, X-Peer-Kind headers in bridge context middleware
- Add BridgeContext struct to bundle per-call context for MCP config generation
- Include channel routing headers in per-session MCP config files
- Expose "message" tool via MCP bridge for cross-channel messaging
- Add extract helpers for new option keys in claude_cli_session.go
* feat(mcp-bridge): forward media attachments to outbound message bus
- Wire MessageBus into gateway server and MCP bridge handler
- Publish tool result media files to outbound bus for channel delivery
- Extract channel/chatID/peerKind from tool context for proper routing
- Add mimeFromExt helper for content-type detection on attachments
* feat(mcp-bridge): inject per-agent DB-backed MCP servers into Claude CLI config
- Add MCPServerLookup type to resolve agent-specific MCP servers from DB
- Wire MCPServerStore through provider registration and HTTP handler
- Extract mcpServerEntryToConfig helper to deduplicate transport config logic
- Add JSON-to-Go helpers (jsonToStringSlice, jsonToStringMap) for DB fields
- Merge per-agent MCP servers at config write time without overriding static entries
* fix(mcp-bridge): use Media struct fields and prefer explicit MimeType
- Map Media.Path to attachment URL instead of treating Media as string
- Use Media.MimeType when available, fall back to extension-based detection
* refactor(providers): deduplicate option extractors and extract bridge media forwarding
- Replace per-field extractors (extractSessionKey, extractAgentID, etc.) with generic extractStringOpt/extractBoolOpt
- Add bridgeContextFromOpts helper to build BridgeContext in one call
- Extract forwardMediaToOutbound from inline block in makeToolHandler
- Change NewBridgeServer msgBus param from variadic to explicit pointer
* fix(providers): validate provider_type on update instead of silently dropping it
- Add explicit validation against ValidProviderTypes with 400 response
- Remove silent delete(updates, "provider_type") that hid invalid values
- Caller now receives clear error when submitting unsupported provider_type
* fix(providers): add header injection validation to MCP bridge headers
- Extend CRLF/null-byte checks to agentID, channel, chatID, and peerKind
- Previously only userID had header injection prevention
- Prevents HTTP header injection via crafted values in MCP config
* fix(mcp-bridge): sign all context fields in HMAC and remove legacy code
- Sign all 5 bridge context fields (agentID|userID|channel|chatID|peerKind)
in HMAC instead of only agentID|userID to prevent channel routing forgery
- Propagate context.Context into MCPServerLookup to respect request
cancellation instead of using context.Background()
- Remove legacy BuildCLIMCPConfig, WithClaudeCLIMCPConfig, mcpConfigPath,
and mcpCleanup (dead code since system is PG-only)
- Use mime.TypeByExtension before custom fallback in mimeFromExt
- Add debug log when media forwarding is skipped due to missing context
- Add thread-safety comment to SetMCPServerLookup
---------
Co-authored-by: Nam Nguyen Ngoc <namnn.0911@gmail.com>
Co-authored-by: viettranx <viettranx@gmail.com>
Fixes#87. crypto.randomUUID() is unavailable in browsers older than
Chrome 92 / Firefox 95 / Safari 15.4, causing an uncaught TypeError
that crashes the React tree and renders a blank page when opening
builtin-tool settings or MCP grant dialogs.
Add a uniqueId() helper that uses crypto.randomUUID() when available
and falls back to a Math.random-based UUID v4 generator otherwise.
Co-authored-by: Luvu182 <208665161+Luvu182@users.noreply.github.com>
viettranx