1. Split the oversized `cmd/agent` entrypoint into a multi-file package and update dev/build scripts to use the package path instead of compiling `main.go` directly.
2. Add a new `memoh` terminal UI for local bot chat, with Bubble Tea
Provider preset YAML files under conf/providers were not bundled into
the server Docker image or preserved by the install script, so fresh
deployments started without any pre-configured LLM providers.
Replace the single browser image (which required local build) with three
prebuilt images: browser-chromium, browser-firefox, and browser (both).
Each is exposed as a separate Docker Compose profile so users can simply
`docker compose --profile browser-chromium up -d` without any local build
step, significantly reducing install time.
* feat(web): add provider oauth management ui
* feat: add OAuth callback support on port 1455
* feat: enhance reasoning effort options and support for OpenAI Codex OAuth
* feat: update twilight-ai dependency to v0.3.4
* refactor: promote openai-codex to first-class client_type, remove auth_type
Replace the previous openai-responses + metadata auth_type=openai-codex-oauth
combo with a dedicated openai-codex client_type. OAuth requirement is now
determined solely by client_type, eliminating the auth_type concept from the
LLM provider domain entirely.
- Add openai-codex to DB CHECK constraint (migration 0047) with data migration
- Add ClientTypeOpenAICodex constant and dedicated SDK/probe branches
- Remove AuthType from SDKModelConfig, ModelCredentials, TriggerConfig, etc.
- Simplify supportsOAuth to check client_type == openai-codex
- Add conf/providers/codex.yaml preset with Codex catalog models
- Frontend: replace auth_type selector with client_type-driven OAuth UI
---------
Co-authored-by: Acbox <acbox0328@gmail.com>
* refactor(agent): replace TypeScript agent gateway with in-process Go agent using twilight-ai SDK
- Remove apps/agent (Bun/Elysia gateway), packages/agent (@memoh/agent),
internal/bun runtime manager, and all embedded agent/bun assets
- Add internal/agent package powered by twilight-ai SDK for LLM calls,
tool execution, streaming, sential logic, tag extraction, and prompts
- Integrate ToolGatewayService in-process for both built-in and user MCP
tools, eliminating HTTP round-trips to the old gateway
- Update resolver to convert between sdk.Message and ModelMessage at the
boundary (resolver_messages.go), keeping agent package free of
persistence concerns
- Prepend user message before storeRound since SDK only returns output
messages (assistant + tool)
- Clean up all Docker configs, TOML configs, nginx proxy, Dockerfile.agent,
and Go config structs related to the removed agent gateway
- Update cmd/agent and cmd/memoh entry points with setter-based
ToolGateway injection to avoid FX dependency cycles
* fix(web): move form declaration before computed properties that reference it
The `form` reactive object was declared after computed properties like
`selectedMemoryProvider` and `isSelectedMemoryProviderPersisted` that
reference it, causing a TDZ ReferenceError during setup.
* fix: prevent UTF-8 character corruption in streaming text output
StreamTagExtractor.Push() used byte-level string slicing to hold back
buffer tails for tag detection, which could split multi-byte UTF-8
characters. After json.Marshal replaced invalid bytes with U+FFFD,
the corruption became permanent — causing garbled CJK characters (�)
in agent responses.
Add safeUTF8SplitIndex() to back up split points to valid character
boundaries. Also fix byte-level truncation in command/formatter.go
and command/fs.go to use rune-aware slicing.
* fix: add agent error logging and fix Gemini tool schema validation
- Log agent stream errors in both SSE and WebSocket paths with bot/model context
- Fix send tool `attachments` parameter: empty `items` schema rejected by
Google Gemini API (INVALID_ARGUMENT), now specifies `{"type": "string"}`
- Upgrade twilight-ai to d898f0b (includes raw body in API error messages)
* chore(ci): remove agent gateway from Docker build and release pipelines
Agent gateway has been replaced by in-process Go agent; remove the
obsolete Docker image matrix entry, Bun/UPX CI steps, and agent-binary
build logic from the release script.
* fix: preserve attachment filename, metadata, and container path through persistence
- Add `name` column to `bot_history_message_assets` (migration 0034) to
persist original filenames across page refreshes.
- Add `metadata` JSONB column (migration 0035) to store source_path,
source_url, and other context alongside each asset.
- Update SQL queries, sqlc-generated code, and all Go types (MessageAsset,
AssetRef, OutboundAssetRef, FileAttachment) to carry name and metadata
through the full lifecycle.
- Extract filenames from path/URL in AttachmentsResolver before clearing
raw paths; enrich streaming event metadata with name, source_path, and
source_url in both the WebSocket and channel inbound ingestion paths.
- Implement `LinkAssets` on message service and `LinkOutboundAssets` on
flow resolver so WebSocket-streamed bot attachments are persisted to the
correct assistant message after streaming completes.
- Frontend: update MessageAsset type with metadata field, pass metadata
through to attachment items, and reorder attachment-block.vue template
so container files (identified by metadata.source_path) open in the
sidebar file manager instead of triggering a download.
* refactor(agent): decouple built-in tools from MCP, load via ToolProvider interface
Migrate all 13 built-in tool providers from internal/mcp/providers/ to
internal/agent/tools/ using the twilight-ai sdk.Tool structure. The agent
now loads tools through a ToolProvider interface instead of the MCP
ToolGatewayService, which is simplified to only manage external federation
sources. This enables selective tool loading and removes the coupling
between business tools and the MCP protocol layer.
* refactor(flow): split monolithic resolver.go into focused modules
Break the 1959-line resolver.go into 12 files organized by concern:
- resolver.go: core orchestration (Resolver struct, resolve, Chat, prepareRunConfig)
- resolver_stream.go: streaming (StreamChat, StreamChatWS, tryStoreStream)
- resolver_trigger.go: schedule/heartbeat triggers
- resolver_attachments.go: attachment routing, inlining, encoding
- resolver_history.go: message loading, deduplication, token trimming
- resolver_store.go: persistence (storeRound, storeMessages, asset linking)
- resolver_memory.go: memory provider integration
- resolver_model_selection.go: model selection and candidate matching
- resolver_identity.go: display name and channel identity resolution
- resolver_settings.go: bot settings, loop detection, inbox
- user_header.go: YAML front-matter formatting
- resolver_util.go: shared utilities (sanitize, normalize, dedup, UUID)
* fix(agent): enable Anthropic extended thinking by passing ReasoningConfig to provider
Anthropic's thinking requires WithThinking() at provider creation time,
unlike OpenAI which uses per-request ReasoningEffort. The config was
never wired through, so Claude models could not trigger thinking.
* refactor(agent): extract prompts into embedded markdown templates
Move inline prompt strings from prompt.go into separate .md files under
internal/agent/prompts/, using {{key}} placeholders and a simple render
engine. Remove obsolete SystemPromptParams fields (Language,
MaxContextLoadTime, Channels, CurrentChannel) and their call-site usage.
* fix: lint
* feat(mcp): workspace container with bridge architecture
Migrate MCP containers to use UDS-based bridge communication instead of
TCP gRPC. Containers now mount runtime binaries and Unix domain sockets
from the host, eliminating the need for a dedicated MCP Docker image.
- Remove Dockerfile.mcp and entrypoint.sh in favor of standard base images
- Add toolkit Dockerfile for building MCP binary separately
- Containers use bind mounts for /opt/memoh (runtime) and /run/memoh (UDS)
- Update all config files with new runtime_path and socket_dir settings
- Support custom base images per bot (debian, alpine, ubuntu, etc.)
- Legacy container detection and TCP fallback for pre-bridge containers
- Frontend: add base image selector in container creation UI
* feat(container): SSE progress bar for container creation
Add real-time progress feedback during container image pull and creation
using Server-Sent Events, without breaking the existing synchronous JSON
API (content negotiation via Accept header).
Backend:
- Add PullProgress/LayerStatus types and OnProgress callback to
PullImageOptions (containerd service layer)
- DefaultService.PullImage polls ContentStore.ListStatuses every 500ms
when OnProgress is set; AppleService ignores it
- CreateContainer handler checks Accept: text/event-stream and switches
to SSE branch: pulling → pull_progress → creating → complete/error
Frontend:
- handleCreateContainer/handleRecreateContainer use fetch + SSE instead
of the SDK's synchronous postBotsByBotIdContainer
- Progress bar shows layer-level pull progress (offset/total) during
pulling phase and indeterminate animation during creating phase
- i18n keys added for pullingImage and creatingContainer (en/zh)
* fix(container): clear stale legacy route and type create SSE
* fix(ci): resolve lint errors and arm64 musl node.js download
- Fix unused-receiver lint: rename `s` to `_` on stub methods in
manager_legacy_test.go
- Fix sloglint: use slog.DiscardHandler instead of
slog.NewTextHandler(io.Discard, nil)
- Handle missing arm64 musl Node.js builds: unofficial-builds.nodejs.org
does not provide arm64 musl binaries, fall back to glibc build
* fix(lint): address errcheck, staticcheck, and gosec findings
- Discard os.Setenv/os.Remove return values explicitly with _
- Use omitted receiver name instead of _ (staticcheck ST1006)
- Tighten directory permissions from 0o755 to 0o750 (gosec G301)
* fix(lint): sanitize socket path to satisfy gosec G703
filepath.Clean the env-sourced socket path before os.Remove
to avoid path-traversal taint warning.
* fix(lint): use nolint directive for gosec G703 on socket path
filepath.Clean does not satisfy gosec's taint analysis. The socket
path comes from MCP_SOCKET_PATH env (operator-configured) or a
compiled-in default, not from end-user input.
* refactor: rename MCP container/bridge to workspace/bridge
Split internal/mcp/ to separate container lifecycle management from
Model Context Protocol connections, eliminating naming confusion:
- internal/mcp/ (container mgmt) → internal/workspace/
- internal/mcp/mcpclient/ → internal/workspace/bridge/
- internal/mcp/mcpcontainer/ → internal/workspace/bridgepb/
- cmd/mcp/ → cmd/bridge/
- config: MCPConfig → WorkspaceConfig, [mcp] → [workspace]
- container prefix: mcp-{id} → workspace-{id}
- labels: mcp.bot_id → memoh.bot_id, add memoh.workspace=v1
- socket: mcp.sock → bridge.sock, env BRIDGE_SOCKET_PATH
- runtime: /opt/memoh/runtime/mcp → /opt/memoh/runtime/bridge
- devenv: mcp-build.sh → bridge-build.sh
Legacy containers (mcp- prefix) detected by container name prefix
and handled via existing fallback path.
* fix(container): use memoh.workspace=v3 label value
* refactor(container): drop LegacyBotLabelKey, infer bot ID from container name
Legacy containers use mcp-{botID} naming, so bot ID can be derived
via TrimPrefix instead of looking up the mcp.bot_id label.
* fix(workspace): resolve containers via manager and drop gateway container ID
* docs: fix stale mcp references in AGENTS.md and DEPLOYMENT.md
* refactor(workspace): move container lifecycle ownership into manager
* dev: isolate local devenv from prod config
* toolkit: support musl node runtime
* containerd: fix fallback resolv.conf permissions
* web: preserve container create progress on completion
* web: add bot creation wait hint
* fix(workspace): preserve image selection across recreate
* feat(web): shorten default docker hub image refs
* fix(container): address code review findings
- Remove synchronous CreateContainer path (SSE-only now)
- Move flusher check before WriteHeader to avoid committed 200 on error
- Fix legacy container IP not cached via ensureContainerAndTask path
- Add atomic guard to prevent stale pull_progress after PullImage returns
- Defensive copy for tzEnv slice to avoid mutating shared backing array
- Restore network failure severity in restartContainer (return + Error)
- Extract duplicate progress bar into ContainerCreateProgress component
- Fix codesync comments to use repo-relative paths
- Add SaaS image validation note and kernel version comment on reaper
* refactor(devenv): extract toolkit install into shared script
Unify the Node.js + uv download logic into docker/toolkit/install.sh,
used by the production Dockerfile and runnable locally for dev.
Dev environment no longer bakes toolkit into the Docker image — it is
volume-mounted from .toolkit/ instead, so wrapper script changes take
effect immediately without rebuilding. The entrypoint checks for the
toolkit directory and prints a clear error if missing.
* fix(ci): address go ci failures
* chore(docker): remove unused containerd image
* refactor(config): rename workspace image key
* fix(workspace): fix legacy container data loss on migration and stop swallowing errors
Three root causes were identified and fixed:
1. Delete() used hardcoded "workspace-" prefix to look up legacy "mcp-"
containers, causing GetContainer to return NotFound. CleanupBotContainer
then silently skipped the error and deleted the DB record without ever
calling PreserveData. Fix: resolve the actual container ID via
ContainerID() (DB → label → scan) before operating.
2. Multiple restore error paths were silently swallowed (logged as Warn
but not returned), so the user saw HTTP 200/204 with no data and no
error. Fix: all errors in the preserve/restore chain now block the
workflow and propagate to the caller.
3. tarGzDir used cached DirEntry.Info() for tar header size, which on
overlayfs can differ from the actual file size, causing "archive/tar:
write too long". Fix: open the file first, Fstat the fd for a
race-free size, and use LimitReader as a safeguard.
Also adds a "restoring" SSE phase so the frontend shows a progress
indicator ("Restoring data, this may take a while...") during data
migration on container recreation.
* refactor(workspace): single-point container ID resolution
Replace the `containerID func(string) string` field with a single
`resolveContainerID(ctx, botID)` method that resolves the actual
container ID via DB → label → scan → fallback. All ~16 lookup
callsites across manager.go, dataio.go, versioning.go, and
manager_lifecycle.go now go through this single resolver, which
correctly handles both legacy "mcp-" and new "workspace-" containers.
Only `ensureBotWithImage` inlines `ContainerPrefix + botID` for
creating brand-new containers — every other path resolves dynamically.
* fix(web): show progress during data backup phase of container recreate
The recreate flow (delete with preserve_data + create with restore_data)
blocked on the DELETE call while backing up /data with no progress
indication. Add a 'preserving' phase to the progress component so
users see "正在备份数据..." instead of an unexplained hang.
* chore: remove [MYDEBUG] debug logging
Clean up all 112 temporary debug log statements added during the
legacy container migration investigation. Kept only meaningful
warn-level logs for non-fatal errors (network teardown, rename
failures).
* refactor(browser): split browser cores via build ARG, add core selector
- Replace playwright official image with ubuntu:noble base in both
docker/Dockerfile.browser and devenv/Dockerfile.browser; install
browsers at build time driven by ARG/ENV BROWSER_CORES
- Add GET /cores endpoint to Browser Gateway reporting available cores
- Proxy GET /browser-contexts/cores in Go handler to Browser Gateway
- Add `core` field to BrowserContextConfigModel and GatewayBrowserContext;
context creation selects the appropriate browser instance by core
- Frontend context-setting page fetches available cores and renders a
core selector; saves core as part of the config JSON
- install.sh prompts for browser core selection and writes BROWSER_CORES
to .env; builds the browser image locally before docker compose up
- Regenerate OpenAPI spec and TypeScript SDK
* fix: lint
MCP containers spawning child processes (npx, uvx, etc.) left zombies
because the Go binary running as PID 1 does not reap orphaned children.
Add dumb-init as the entrypoint init process.
* fix(containerd): prevent silent network failures from leaving containers unreachable
Container network setup failures were silently swallowed at multiple
points in the call chain, leaving containers in a "running but
unreachable" ghost state. This patch closes every silent-failure path:
- setupCNINetwork: return error when CNI yields no usable IP
- Manager.Start: roll back container when IP is empty instead of
returning success
- ensureContainerAndTask: extract setupNetworkOrFail with 1 retry,
propagate error to callers
- ReconcileContainers: stop reporting "healthy" when network setup fails
- recoverContainerIP: retry up to 2 times with backoff for transient
CNI failures (IPAM lock contention, etc.)
- gRPC Pool: evict connections stuck in Connecting state for >30s
* fix(containerd): clean stale cni0 bridge on startup to prevent MAC error
After a Docker container restart, the cni0 bridge interface can linger
with a zeroed MAC (00:00:00:00:00:00) and DOWN state. The CNI bridge
plugin then fails with "could not set bridge's mac: invalid argument",
making all MCP containers unreachable.
Two-layer fix:
- Entrypoint: delete cni0 and flush IPAM state before starting containerd
- Go: detect bridge MAC errors in setupCNINetwork and auto-delete cni0
before retrying, as defense-in-depth for runtime recovery
* fix(containerd): use exec.CommandContext to satisfy noctx linter
jsdom relies on Node.js-specific APIs that Bun cannot properly resolve
when running a bundled artifact. This caused an EISDIR error in Docker
containers (Bun tried to read the jsdom directory as a file).
Replace jsdom with linkedom, a lightweight pure-JS DOM implementation
that is fully compatible with Bun and @mozilla/readability. Also remove
the --external jsdom build flag since linkedom bundles cleanly.
Closes#181
Replace the host bind-mount + containerd exec approach with a per-bot
in-container gRPC server (ContainerService, port 9090). All file I/O,
exec, and MCP stdio sessions now go through gRPC instead of running
shell commands or reading host-mounted directories.
Architecture changes:
- cmd/mcp: rewritten as a gRPC server (ContainerService) with full
file and exec API (ReadFile, WriteFile, ListDir, ReadRaw, WriteRaw,
Exec, Stat, Mkdir, Rename, DeleteFile)
- internal/mcp/mcpcontainer: protobuf definitions and generated stubs
- internal/mcp/mcpclient: gRPC client wrapper with connection pool
(Pool) and Provider interface for dependency injection
- mcp.Manager: add per-bot IP cache, gRPC connection pool, and
SetContainerIP/MCPClient methods; remove DataDir/Exec helpers
- containerd.Service: remove ExecTask/ExecTaskStreaming; network setup
now returns NetworkResult{IP} for pool routing
- internal/fs/service.go: deleted (replaced by mcpclient)
- handlers/fs.go: deleted; MCP stdio session logic moved to mcp_stdio.go
- container provider Executor: all tools (read/write/list/edit/exec)
now call gRPC client instead of running shell via exec
- storefs, containerfs, media, skills, memory: all I/O ported to
mcpclient.Provider
Database:
- migration 0022: drop host_path column from containers table
One-time data migration:
- migrateBindMountData: on first Start() after upgrade, copies old
bind-mount data into the container via gRPC, then renames src dir
to prevent re-migration; runs in background goroutine
Bug fixes:
- mcp_stdio: callRaw now returns full JSON-RPC envelope
{"jsonrpc","id","result"|"error"} matching protocol spec;
explicit "initialize" call now advances session init state to
prevent duplicate handshake on next non-initialize call
- mcpclient Pool: properly evict stale gRPC connection after snapshot
replace (container process recreated); use SetContainerIP instead
of direct map write so IP changes always evict pool entry
- migrateBindMountData: walkErr on directories now counted as failure
so partially-walked trees don't get incorrectly marked as migrated
- cmd/mcp/Dockerfile: removed dead file (docker/Dockerfile.mcp is the
canonical production build)
Tests:
- provider_test.go: restored with bufconn in-process gRPC mock
(fakeContainerService + staticProvider), 14 cases covering all 5
tools plus edge cases
- mcp_session_test.go: new, covers JSON-RPC envelope, init state
machine, pending cleanup on cancel/close, readLoop cancel
- storefs/service_test.go: restored (pure function roundtrip tests)
cni.Remove() failure on stale iptables state blocked the retry
cni.Setup(), leaving bot containers without SNAT/MASQUERADE.
- Ignore cni.Remove() error so retry Setup always runs
- Add global MASQUERADE rule in entrypoints as belt-and-suspenders
Closes#161
Add mcp-build.sh that compiles the MCP binary and packages it as an
OCI image layer on top of the base rootfs, imported directly into
containerd. Air triggers rebuild on code changes, cleaning stale
containers automatically.
Consolidate dev-only files (Dockerfiles, entrypoint, config, build
script) into devenv/ to separate dev tooling from production artifacts.
Skip image pull when already imported to speed up dev startup.
* feat(devenv): add containerized development environment
Replace local-process dev workflow with a fully containerized stack
using docker compose. This enables consistent development across
machines without requiring local Go/Node toolchains or containerd.
- Add Dockerfile.server.dev with containerd + CNI networking support
- Add Dockerfile.web.dev for frontend dev server
- Add server-dev-entrypoint.sh for containerd lifecycle management
- Expand devenv/docker-compose.yml with server, agent, web, migrate
and deps services with proper health checks and dependency ordering
- Update app.dev.toml to use container service names instead of localhost
- Refactor mise.toml dev tasks to drive docker compose workflow
- Support agent_gateway.server_addr in config package for inter-container
communication
* feat(devenv): add hot-reload and registry mirror support
- Add air for Go server hot-reload in dev containers
- Fix agent_gateway host in dev config (0.0.0.0 -> agent)
- Add configurable registry mirror for China mainland users
- Unify MCP image refs via MCPConfig.ImageRef()
* feat(scripts): add China mainland mirror option to install script
Prompt users to opt-in to memoh.cn mirror during installation,
which applies docker-compose.cn.yml overlay and sets registry
in config.toml for MCP image pulls.
- Add pid: host to containerd service so server can access MCP container
network namespaces via /proc/PID/ns/net for CNI setup
- Add Node.js, npm, Python 3, uv to embedded MCP image rootfs so users
can run npx/uvx MCP servers inside containers
晨苒