Bot
f3e25f7503
fix: push type filters into SQL WHERE clause + expose in MCP search
...
Type filtering (include_types/exclude_types) was applied post-fetch after
the database LIMIT, which could return zero results when all top-N
results were of the filtered type. This pushes type filters into the SQL
WHERE clause in keyword_search_with_scores() so they apply before the
limit. Semantic results still get post-fetch filtering as a safety net
since the vector index cannot filter by type.
Also adds hybrid_search_filtered() as the new primary method, with the
original hybrid_search() delegating to it with no filters for backward
compatibility. The MCP search tool now exposes include_types and
exclude_types parameters.
Includes 5 new test cases covering include, exclude, precedence,
empty results, and backward compatibility.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-13 14:42:16 -05:00
Sam Valladares
04781a95e2
feat: v2.0.4 "Deep Reference" — cognitive reasoning engine + 10 bug fixes
...
New features:
- deep_reference tool (#22 ): 8-stage cognitive reasoning pipeline with FSRS-6
trust scoring, intent classification (FactCheck/Timeline/RootCause/Comparison/
Synthesis), spreading activation expansion, temporal supersession, trust-weighted
contradiction analysis, relation assessment, dream insight integration, and
algorithmic reasoning chain generation — all without calling an LLM
- cross_reference (#23 ): backward-compatible alias for deep_reference
- retrieval_mode parameter on search (precise/balanced/exhaustive)
- get_batch action on memory tool (up to 20 IDs per call)
- Token budget raised from 10K to 100K on search + session_context
- Dates (createdAt/updatedAt) on all search results and session_context lines
Bug fixes (GitHub Issue #25 — all 10 resolved):
- state_transitions empty: wired record_memory_access into strengthen_batch
- chain/bridges no storage fallback: added with edge deduplication
- knowledge_edges dead schema: documented as deprecated
- insights not persisted from dream: wired save_insight after generation
- find_duplicates threshold dropped: serde alias fix
- search min_retention ignored: serde aliases for snake_case params
- intention time triggers null: removed dead trigger_at embedding
- changelog missing dreams: added get_dream_history + event integration
- phantom Related IDs: clarified message text
- fsrs_cards empty: documented as harmless dead schema
Security hardening:
- HTTP transport CORS: permissive() → localhost-only
- Auth token panic guard: &token[..8] → safe min(8) slice
- UTF-8 boundary fix: floor_char_boundary on content truncation
- All unwrap() removed from HTTP transport (unwrap_or_else fallback)
- Dream memory_count capped at 500 (prevents O(N²) hang)
- Dormant state threshold aligned (0.3 → 0.4)
Stats: 23 tools, 758 tests, 0 failures, 0 warnings, 0 unwraps in production
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-09 16:15:26 -05:00
Sam Valladares
c2d28f3433
feat: Vestige v2.0.0 "Cognitive Leap" — 3D dashboard, HyDE search, WebSocket events
...
The biggest release in Vestige history. Complete visual and cognitive overhaul.
Dashboard:
- SvelteKit 2 + Three.js 3D neural visualization at localhost:3927/dashboard
- 7 interactive pages: Graph, Memories, Timeline, Feed, Explore, Intentions, Stats
- WebSocket event bus with 16 event types, real-time 3D animations
- Bloom post-processing, GPU instanced rendering, force-directed layout
- Dream visualization mode, FSRS retention curves, command palette (Cmd+K)
- Keyboard shortcuts, responsive mobile layout, PWA installable
- Single binary deployment via include_dir! (22MB)
Engine:
- HyDE query expansion (intent classification + 3-5 semantic variants + centroid)
- fastembed 5.11 with optional Nomic v2 MoE + Qwen3 reranker + Metal GPU
- Emotional memory module (#29 )
- Criterion benchmark suite
Backend:
- Axum WebSocket at /ws with heartbeat + event broadcast
- 7 new REST endpoints for cognitive operations
- Event emission from MCP tools via shared broadcast channel
- CORS for SvelteKit dev mode
Distribution:
- GitHub issue templates (bug report, feature request)
- CHANGELOG with comprehensive v2.0 release notes
- README updated with dashboard docs, architecture diagram, comparison table
734 tests passing, zero warnings, 22MB release binary.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-22 03:07:25 -06:00
Sam Valladares
5b90a73055
feat: Vestige v1.9.1 AUTONOMIC — self-regulating memory with graph visualization
...
Retention Target System: auto-GC low-retention memories during consolidation
(VESTIGE_RETENTION_TARGET env var, default 0.8). Auto-Promote: memories
accessed 3+ times in 24h get frequency-dependent potentiation. Waking SWR
Tagging: promoted memories get preferential 70/30 dream replay. Improved
Consolidation Scheduler: triggers on 6h staleness or 2h active use.
New tools: memory_health (retention dashboard with distribution buckets,
trend tracking, recommendations) and memory_graph (subgraph export with
Fruchterman-Reingold force-directed layout, up to 200 nodes).
Dream connections now persist to database via save_connection(), enabling
memory_graph traversal. Schema Migration V8 adds waking_tag, utility_score,
times_retrieved/useful columns and retention_snapshots table. 21 MCP tools.
v1.9.1 fixes: ConnectionRecord export, UTF-8 safe truncation, link_type
normalization, utility_score clamping, only-new-connections persistence,
70/30 split capacity fill, nonexistent center_id error handling.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-21 02:02:06 -06:00
Sam Valladares
ce520bb246
chore: license AGPL-3.0, zero clippy warnings, CHANGELOG through v1.6.0
...
License:
- Replace MIT/Apache-2.0 with AGPL-3.0-only across all crates and npm packages
- Replace LICENSE file with official GNU AGPL-3.0 text
- Remove LICENSE-MIT and LICENSE-APACHE
Code quality:
- Fix all 44 clippy warnings (zero remaining)
- Collapsible if statements, redundant closures, manual Option::map
- Remove duplicate #[allow(dead_code)] attributes in deprecated tool modules
- Add Default impl for CognitiveEngine
- Replace manual sort_by with sort_by_key
Documentation:
- Update CHANGELOG with v1.2.0, v1.3.0, v1.5.0, v1.6.0 entries
- Update README with v1.6.0 highlights and accurate stats (52K lines, 1100+ tests)
- Add fastembed-rs/ to .gitignore
- Add fastembed-rs to workspace exclude
1115 tests passing, zero warnings, RUSTFLAGS="-Dwarnings" clean.
2026-02-19 03:00:39 -06:00
Sam Valladares
495a88331f
feat: Vestige v1.6.0 — 6x storage reduction, neural reranking, instant startup
...
Four internal optimizations for dramatically better performance:
1. F16 vector quantization (ScalarKind::F16 in USearch) — 2x storage savings
2. Matryoshka 256-dim truncation (768→256) — 3x embedding storage savings
3. Convex Combination fusion (0.3 keyword / 0.7 semantic) replacing RRF
4. Cross-encoder reranker (Jina Reranker v1 Turbo via fastembed TextRerank)
Combined: 6x vector storage reduction, ~20% better retrieval quality.
Cross-encoder loads in background — server starts instantly.
Old 768-dim embeddings auto-migrated on load.
614 tests pass, zero warnings.
2026-02-19 01:09:39 -06:00
Sam Valladares
927f41c3e4
feat: Vestige v1.5.0 — Cognitive Engine, memory dreaming, graph exploration, predictive retrieval
...
28-module CognitiveEngine with full neuroscience pipeline on every tool call.
FSRS-6 now fully automatic: periodic consolidation (6h timer + inline every
100 tool calls), real retrievability formula, episodic-to-semantic auto-merge,
cross-memory reinforcement, Park et al. triple retrieval scoring, ACT-R
base-level activation, personalized w20 optimization.
New tools (19 → 23):
- dream: memory consolidation via replay, discovers hidden connections
- explore_connections: graph traversal (chain, associations, bridges)
- predict: proactive retrieval based on context and activity patterns
- restore: memory restore from JSON backups
All existing tools upgraded with cognitive pre/post processing pipelines.
33 files changed, ~4,100 lines added.
2026-02-18 23:34:15 -06:00
Sam Valladares
34f5e8d52a
feat: Vestige v1.2.0 — dashboard, temporal tools, maintenance tools, detail levels
...
Add web dashboard (axum) on port 3927 with memory browser, search, and
system stats. New MCP tools: memory_timeline, memory_changelog,
health_check, consolidate, stats, backup, export, gc. Search now supports
detail_level (brief/summary/full) to control token usage. Add backup_to()
and get_recent_state_transitions() to storage layer. Bump to v1.2.0.
2026-02-12 04:33:05 -06:00
Sam Valladares
8bb6500985
feat(v1.1): consolidate 29 tools → 8 unified tools + CLI
...
Tool Consolidation:
- search: merges recall, semantic_search, hybrid_search
- memory: merges get_knowledge, delete_knowledge, get_memory_state
- codebase: merges remember_pattern, remember_decision, get_codebase_context
- intention: merges all 5 intention tools into action-based API
New CLI Binary:
- vestige stats [--tagging] [--states]
- vestige health
- vestige consolidate
- vestige restore <file>
Documentation:
- Verify all neuroscience claims against codebase
- Fix Memory States table: "Retention" → "Accessibility" with formula
- Clarify Spreading Activation: embedding similarity vs full network module
- Update Synaptic Tagging: clarify 9h/2h implementation vs biology
- Add comprehensive FAQ with 30+ questions
- Add storage modes: global, per-project, multi-Claude household
- Add CLAUDE.md setup instructions
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2026-01-26 01:31:58 -06:00
