hugo/RuDS
1
0
Fork 0
Code Issues 1 Pull Requests Actions Packages Projects Releases Wiki Activity

Implement persistent conversational AI chat

Browse Source
This commit is contained in:
Chili Palmer
2026-07-19 16:18:01 +02:00
parent fb5cae2131
commit ac611e3f7f
31 changed files with 4035 additions and 43 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
RUST_PLAN_EXTENSION.md
View File

@@ -31,17 +31,17 @@ Done:
- Localized native import sidebar/editor with WXR and uploads pickers, cached analysis reopening, conflict resolution, manual and airplane-gated AI taxonomy mapping, item review, live progress/ETA, and execution results.
- Taxonomy/posts/media/pages execution through core persistence engines in recoverable 500-item batches, including filesystem rollback, source metadata/status/timestamps, unique-slug import, overwrite/ignore behavior, and media-parent links.
### Conversational AI and agent tools — Open
### Conversational AI and agent tools — Complete
Open:
Done:
- Conversation persistence and chat UI.
- Streaming OpenAI-compatible responses and tool-call parsing.
- Tools over posts, media, templates, search, and other core engines.
- Agent integrations such as Claude Code and Copilot where required by the specs.
- Replace the current Chat placeholders with the working feature.
- Persistent conversation/message repositories with rename, reopen, deletion, model and provider-session selection, and four-way token accounting.
- OpenAI-compatible SSE streaming with split-frame content/tool assembly, provider-error handling, independent cancellation, bounded tool rounds, and context truncation that preserves system messages and tool pairs.
- Project-aware tools over statistics, FTS search, posts, media, templates, scripts, tags/categories, metadata mutation through shared engines, and allowlisted workspace navigation.
- Localized Chat sidebar/editor with conversation and model controls, safe GFM text rendering with blocked external images, streaming/tool state, multiline send/stop controls, and status-bar token totals.
- Online/airplane endpoint routing uses the shared secure endpoint and model infrastructure; unavailable modes direct the user to the existing localized AI settings.
Core endpoint settings, offline gating, key storage, model discovery, and seven one-shot operations are already implemented.
Persistent A2UI surfaces remain separately tracked below.
### Embeddings, semantic search, and duplicates — Open