bb6dbc6313
chore: update go module dependencies
2026-04-19 22:20:46 +09:00
01dd41eca7
feat: enhance speech provider functionality with advanced settings and model import capabilities
2026-04-19 22:01:05 +09:00
e1b94e1d41
Merge branch 'main' into feat/speech-support
2026-04-19 21:56:34 +09:00
030bc8817f
chore: remove go.mod replace
2026-04-19 21:50:29 +09:00
460a0307f3
feat: enhance speech provider functionality with advanced settings and model import capabilities
2026-04-19 21:10:22 +09:00
8e013ad1ad
feat(platform): add slack platform support ( #385 )
...
* feat(platform): add slack platform support
* docs: add slack channel setup guide
* feat: normalize slack unicode reactions
* chore(docs): remove unsupport feature
* fix(slack): harden adapter stream and identity handling
- ignore reaction and speech stream events in Slack outbound
streams
- normalize Slack conversation types to framework-standard values
- route DiscoverSelf through the adapter API factory
- add config-scoped Slack user display-name caching
- expand adapter interface assertions and add regression coverage
- add ChannelTypeSlack to well-known channel constants
2026-04-19 14:17:05 +08:00
30ab6b4199
feat: except *-speech client type in llm provider
2026-04-19 14:16:01 +08:00
66808f868b
feat: add SVG support for Deepgram and Elevenlabs with Vue components
2026-04-19 01:08:04 +09:00
9e237e63d4
feat: add icon support for speech providers and update related configurations
2026-04-19 00:19:49 +09:00
f0a0706d75
feat: expand speech provider support with new client types and configuration schema
2026-04-18 20:43:36 +09:00
db777b98ac
fix(agent): stream loop abort, mid-stream retry parity, collector cleanup ( #376 )
...
* fix(agent): align stream retry abort and event collection
* fix(agent): cancel stream on loop detect, harden retry and tool events
* fix(agent): drain previous stream before retry
* fix(lint): ctx ci lint
---------
Co-authored-by: 晨苒 <16112591+chen-ran@users.noreply.github.com >
2026-04-18 03:19:58 +08:00
b534248e19
fix: resolve context switch failure in browser automation ( #380 )
...
* fix: resolve context switch failure in browser automation
* feat: update logo and optimize sidebar empty state
---------
Co-authored-by: 晨苒 <16112591+chen-ran@users.noreply.github.com >
2026-04-18 03:07:17 +08:00
28d11713db
fix: preserve empty skill discovery roots ( #384 )
2026-04-17 23:17:04 +08:00
dee82177d3
feat: add bot-level skill paths configuration ( #383 )
2026-04-17 16:05:00 +08:00
7aa6ec6ca9
chore(ci): add release.yml workflow_dispatch
2026-04-16 18:38:02 +08:00
d3a820b2dc
release: v0.7.0
2026-04-16 18:16:11 +08:00
2f5bb97ab4
chore: update logo to web and documentation
2026-04-16 18:14:18 +08:00
74f04427af
chore(ci): unify release pipeline
2026-04-16 17:58:23 +08:00
68b1efa343
docs: update to v0.7
2026-04-16 17:45:56 +08:00
fc1ee59dea
Revert "release: v0.7.0"
...
This reverts commit 9e3fae7e9b
2026-04-16 17:37:59 +08:00
9e3fae7e9b
release: v0.7.0
2026-04-16 16:50:58 +08:00
9f10033f63
feat(agent): include platform self identities in system prompts
...
Give bots their known per-channel account handles in the system prompt so they can reason about platform-specific self references consistently. Reuse persisted channel self_identity data across chat, discuss, schedule, heartbeat, and subagent prompts.
2026-04-16 16:44:29 +08:00
e0fc2f514e
fix(agent): surface tool calls before input completes
...
Emit tool-call placeholders as soon as tool input streaming starts so long writes appear immediately in chat. Reuse the same UI tool message when full input arrives to avoid duplicate cards, and keep the hook-required test suite green.
2026-04-16 16:42:07 +08:00
1a5b1d6086
fix(web): show send tool calls in chat
...
Remove the frontend filter that hid send tool blocks delivered to the current conversation so assistant actions remain visible in chat.
2026-04-16 16:39:57 +08:00
33e18e7e64
feat(skills): add effective skill resolution and actions ( #377 )
...
* feat(skills): add effective skill resolution and actions
* refactor(workspace): normalize skill-related env and prompt
* chore(api): regenerate skills OpenAPI and SDK artifacts
* feat(web): surface effective skill state in console
* test(skills): cover API and runtime effective state
* fix(web): show adopt action for discovered skills
* fix(web): align skill header and show stateful visibility icon
* refactor(web): compact skill metadata on narrow layouts
* fix(web): constrain long skill text in cards
* refactor(skills): narrow default discovery roots
* fix(skills): harden managed skill path validation
* feat: add path in the results of `use_skill`
---------
Co-authored-by: Acbox <acbox0328@gmail.com >
2026-04-16 13:50:39 +08:00
8e1ed3683f
feat(agent): relax provider http timeouts ( #348 )
2026-04-15 00:07:41 +08:00
38ac907361
chore(docs): update new memoh logo
2026-04-14 22:41:54 +08:00
7cc7b959e2
fix: resolve bot styling issue ( #375 )
2026-04-14 22:31:30 +08:00
5382c3cd27
feat(web): add sidebar collapse
2026-04-14 21:58:29 +08:00
19d5cd606d
chore: update AGENTS.md
2026-04-14 21:56:42 +08:00
2a1c4a2e5f
fix(providers): github copilot icon
2026-04-14 21:30:46 +08:00
27d2b99301
feat: add immediate context compaction API, UI button, and /compact slash command
...
- Add POST /bots/:bot_id/sessions/:session_id/compact endpoint for
synchronous context compaction with fallback to chat model when no
dedicated compaction model is configured
- Add "Compact Now" button to session info panel in the web UI
- Add /compact slash command for triggering compaction from chat
- Regenerate OpenAPI spec and TypeScript SDK
2026-04-14 21:30:05 +08:00
6328281fc2
fix: enforce speech/LLM isolation in providers and models
...
SQL queries (CountProviders, CountModels, ListModels, ListEnabledModels,
ListModelsByProviderID) now exclude speech types. Added IsLLMClientType
guard to prevent cross-domain queries via /models?client_type and
/providers/:id/import-models. Frontend provider forms no longer offer
edge-speech as a client type option.
Also fixed pre-existing SA5011 staticcheck warnings in proxy_test.go
and executor_test.go.
2026-04-14 21:07:27 +08:00
84f1d0612a
refactor: replace context_token_budget with model context_window for context trimming
...
The per-bot context_token_budget column was unused (no frontend UI) and
has been removed. Context trimming now derives the budget from the chat
model's context_window setting, which is already configured per model.
2026-04-14 21:04:42 +08:00
cb44408277
fix(ci): avoid direct secrets access in tauri workflows
2026-04-14 06:51:06 +08:00
4d5f3f9126
release: v0.7.0-beta.4
2026-04-14 06:44:07 +08:00
fa48f59d38
fix(docker): docker-compose.yml typo
2026-04-14 06:43:30 +08:00
0c2527453b
fix(schedule): use anyOf for nullable max_calls
2026-04-14 06:38:49 +08:00
a076b232fc
Merge branch 'pr/350'
2026-04-14 06:37:43 +08:00
1739a7b671
Merge branch 'main' into pr/350
2026-04-14 06:33:19 +08:00
3509947bc0
Fix dangling tool call history
2026-04-14 06:23:48 +08:00
3945bd913d
test(agent): satisfy lint for generate loop tests
2026-04-14 06:23:01 +08:00
33461d7ac1
fix(agent): stop generate loop aborts correctly
2026-04-14 06:23:01 +08:00
59147b255d
fix(agent): guard tool loop state against concurrent tool execution
2026-04-14 06:23:01 +08:00
1a6d12a137
feat(cli): add bots and docker compose commands
2026-04-14 06:22:00 +08:00
d50eeea114
reactor(cli): move memoh cli to tui
...
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
2026-04-14 06:22:00 +08:00
8c9f222783
fix(ui): add acl template tooltip
2026-04-14 05:51:38 +08:00
60517bc2a6
feat(acl): add bot security policy presets
...
Initialize new bots with preset ACL templates and an allow-by-default fallback so common access setups can be selected during bot creation instead of being configured manually afterward.
2026-04-14 05:51:38 +08:00
0e6c8ca451
fix(telegram): dedupe repeated inbound updates
2026-04-14 05:15:45 +08:00
447d647aef
fix(sqlc): cast optional scalar settings args ( #370 )
...
Add explicit scalar casts for optional timezone and context token budget fields so sqlc can parse UpsertBotSettings again. Regenerating sqlc also syncs the stale generated bot/query structs that drifted from the schema.
2026-04-14 04:36:44 +08:00
aki