v0.7.0 - Synthadoc Community Edition

Synthadoc v0.7.0 Community Edition: Key Highlights

Local Web Chat UI

'synthadoc web' CLI starts a local chat interface in your browser, local-only, nothing leaves your machine. Use it the same way you use the CLI: ask questions about your wiki's domain and get streamed, cited answers.

Beyond content queries, the UI also auto-detects the health of your wiki (Explorer, Health Check, or Power User mode) and surfaces context-appropriate prompts: if your wiki has orphan pages or contradictions, those show up as suggested actions rather than generic hints. You can also give operational commands directly in the chat: "run lint", "show wiki status", "schedule ingest every night at 9 PM", the Action Agent parses those and executes them live against your wiki, with results shown inline.

→ Quick-start demo: Step 22 (https://github.com/axoviq-ai/synthadoc/blob/main/docs/user-quick-start-guide.md#step-22--use-the-web-chat-ui)

Streaming Query

Responses now stream token-by-token instead of waiting for the full answer. Works in the CLI, in Obsidian, and in the new web UI. The first token appears almost immediately after you ask; citations and confidence scores appear at the end of the stream once the full answer is assembled. If you need a cached static result, --no-stream still gives you the old behaviour.

→ Quick-start demo: Step 5 (https://github.com/axoviq-ai/synthadoc/blob/main/docs/user-quick-start-guide.md#step-5--query-the-pre-built-wiki-cli--obsidian)

Query Result Caching

Repeated queries return instantly from a local cache instead of hitting the LLM again. The cache is shared across the CLI, the web UI, and the Obsidian plugin, all keyed on the question text and the current wiki state, so it auto-invalidates whenever you ingest new sources or change a page's lifecycle state. synthadoc cache clear drops it when you want a full refresh.

→ Quick-start demo: Step 23 (https://github.com/axoviq-ai/synthadoc/blob/main/docs/user-quick-start-guide.md#step-23--query-caching)

Other enhancements

Also in v0.7.0: expanded built-in hints covering orphan pages, adversarial warnings, and lifecycle state queries; live lint-report and wiki-status summaries available directly from the chat UI.

What's Changed

• fix(export): graphml citation_count now reflects real audit DB counts… by @william-Johnason in #124
• fix(ingest): skip extracted sidecar for web-search intent sources (macOS ENOENT on ':' in filename) & exclude candidates from lifecycle modal, split draft count in status, guard transition by @paulmchen in #125
• feat(schedule): Fix schedule list with cron/next-run/last-run, schedule run/history audit trail by @william-Johnason in #126
• fix: MiniMax reasoning model compatibility + JSON extraction by @william-Johnason in #127
• fix(scaffold): move scaffold LLM call to server side; CLI polls for completion by @paulmchen in #128
• docs: document schedule apply in README and design guide by @william-Johnason in #129
• [v0.7.0] - feat: streaming query, web UI query agent, action agent, live data, query cache by @william-Johnason in #130
• [v0.7.0] fix: streaming truncation, web UI rendering, and action agent improvements by @william-Johnason in #131
• [v0.7.0] feat(action-agent): live lint report, wiki status, and expanded hint coverage by @william-Johnason in #132
• [v0.7.0] release prep by @william-Johnason in #133

Full Changelog: v0.6.0...v0.7.0