Roadmap & status
v0.1.0 substrate baseline was tagged 2026-05-25; v0.1.1 adds the MCP server; v0.1.2 adds the dedicated Wikipedia extractor. The substrate is feature-complete for v0.1.x. Public launch (ToS, Privacy Policy, USPTO TESS, social handles, Show HN, community announcements) is the next horizon.
Works today (v0.1.x)
- All 6 v0.1 source extractors emit typed payloads (X / Reddit / Web / Wikipedia / YouTube / PDF)
- Typed payloads end-to-end: extractor → vault frontmatter (canonical) → SQLite (derived) → recall
- Full Renderer Protocol + Registry (9 renderer implementations) + REST
?format=selector - Media-fetcher registry (HttpxFetcher + opt-in YtDlpFetcher + GalleryDlFetcher)
- Source-tier raw-bytes preservation under a user-configurable
data_root - Bundled MiniLM-L6 semantic recall (zero LLM cost, offline after first fetch)
- Wayback witness (opt-in; off by default — anonymous archive.org SPN is rate-limited)
- P-δ failure handling + visual indicator layers 1 & 2
khiipd validate+khiipd refetch(extraction / media / wayback)- Opt-in video preservation (
[media] download_videos) - MCP server — six tools over stdio for Claude Desktop / Cursor / any MCP client
- Obsidian plugin baseline (capture command + recall sidebar;
types.tsmirrors payloads)
Ahead
- Public launch: PyPI publish, ToS + Privacy Policy, USPTO TESS, social handles, Show HN
- Typed knowledge-graph read/write API — vocabulary + storage schema locked (ADR-0005 / 0007 / 0008); the surface to create + query edges is planned, on no committed version
- Pluggable embedders: local LLM (Ollama), BYOK (OpenAI / Anthropic / Gemini), BM25 fallback (v0.5+)
- Plugin UI rendering of typed-payload + visual indicator + refetch surfaces (v0.5+)
- More sources: Instagram, TikTok, Threads, Bluesky
- Reliable Wayback archiving — authenticated archive.org (SPN2 / S3-key) bring-your-own-credentials tier, so archiving works past the anonymous rate-limit
- Cloud Video Archive premium pillar — managed Source-tier + Whisper transcription (v0.5+)
Tests
847 tests green at v0.1.5. The foundation-vs-on-top discipline (new capability lands in new modules without disturbing existing ones) has held across the build.