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
//! Feedback Tools
2026-01-25 13:30:03 -06:00
//!
//! Promote and demote memories based on outcome quality.
//! Implements preference learning for Vestige.
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
//!
//! v1.5.0: Enhanced with cognitive pipeline:
//! - Reward signal recording (4-channel importance)
//! - Importance tracking (retrieval outcome)
//! - Reconsolidation modification (labile window boost)
//! - Activation network reinforcement
2026-01-25 13:30:03 -06:00
use serde ::Deserialize ;
use serde_json ::Value ;
use std ::sync ::Arc ;
use tokio ::sync ::Mutex ;
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
use crate ::cognitive ::CognitiveEngine ;
use vestige_core ::{ Modification , OutcomeType , Storage } ;
2026-01-25 13:30:03 -06:00
/// Input schema for promote_memory tool
pub fn promote_schema ( ) -> Value {
serde_json ::json! ( {
" type " : " object " ,
" properties " : {
" id " : {
" type " : " string " ,
" description " : " The ID of the memory to promote "
} ,
" reason " : {
" type " : " string " ,
" description " : " Why this memory was helpful (optional, for logging) "
}
} ,
" required " : [ " id " ]
} )
}
/// Input schema for demote_memory tool
pub fn demote_schema ( ) -> Value {
serde_json ::json! ( {
" type " : " object " ,
" properties " : {
" id " : {
" type " : " string " ,
" description " : " The ID of the memory to demote "
} ,
" reason " : {
" type " : " string " ,
" description " : " Why this memory was unhelpful or wrong (optional, for logging) "
}
} ,
" required " : [ " id " ]
} )
}
#[ derive(Debug, Deserialize) ]
struct FeedbackArgs {
id : String ,
reason : Option < String > ,
}
/// Promote a memory (thumbs up) - it led to a good outcome
pub async fn execute_promote (
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
storage : & Arc < Storage > ,
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
cognitive : & Arc < Mutex < CognitiveEngine > > ,
2026-01-25 13:30:03 -06:00
args : Option < Value > ,
) -> Result < Value , String > {
let args : FeedbackArgs = match args {
Some ( v ) = > serde_json ::from_value ( v ) . map_err ( | e | format! ( " Invalid arguments: {} " , e ) ) ? ,
None = > return Err ( " Missing arguments " . to_string ( ) ) ,
} ;
// Validate UUID
uuid ::Uuid ::parse_str ( & args . id ) . map_err ( | _ | " Invalid node ID format " . to_string ( ) ) ? ;
// Get node before for comparison
feat(v2.0.5): Intentional Amnesia — active forgetting via top-down inhibitory control
First AI memory system to model forgetting as a neuroscience-grounded
PROCESS rather than passive decay. Adds the `suppress` MCP tool (#24),
Rac1 cascade worker, migration V10, and dashboard forgetting indicators.
Based on:
- Anderson, Hanslmayr & Quaegebeur (2025), Nat Rev Neurosci — right
lateral PFC as the domain-general inhibitory controller; SIF
compounds with each stopping attempt.
- Cervantes-Sandoval et al. (2020), Front Cell Neurosci PMC7477079 —
Rac1 GTPase as the active synaptic destabilization mechanism.
What's new:
* `suppress` MCP tool — each call compounds `suppression_count` and
subtracts a `0.15 × count` penalty (saturating at 80%) from
retrieval scores during hybrid search. Distinct from delete
(removes) and demote (one-shot).
* Rac1 cascade worker — background sweep piggybacks the 6h
consolidation loop, walks `memory_connections` edges from
recently-suppressed seeds, applies attenuated FSRS decay to
co-activated neighbors. You don't just forget Jake — you fade
the café, the roommate, the birthday.
* 24h labile window — reversible via `suppress({id, reverse: true})`
within 24 hours. Matches Nader reconsolidation semantics.
* Migration V10 — additive-only (`suppression_count`, `suppressed_at`
+ partial indices). All v2.0.x DBs upgrade seamlessly on first launch.
* Dashboard: `ForgettingIndicator.svelte` pulses when suppressions
are active. 3D graph nodes dim to 20% opacity when suppressed.
New WebSocket events: `MemorySuppressed`, `MemoryUnsuppressed`,
`Rac1CascadeSwept`. Heartbeat carries `suppressed_count`.
* Search pipeline: SIF penalty inserted into the accessibility stage
so it stacks on top of passive FSRS decay.
* Tool count bumped 23 → 24. Cognitive modules 29 → 30.
Memories persist — they are INHIBITED, not erased. `memory.get(id)`
returns full content through any number of suppressions. The 24h
labile window is a grace period for regret.
Also fixes issue #31 (dashboard graph view buggy) as a companion UI
bug discovered during the v2.0.5 audit cycle:
* Root cause: node glow `SpriteMaterial` had no `map`, so
`THREE.Sprite` rendered as a solid-coloured 1×1 plane. Additive
blending + `UnrealBloomPass(0.8, 0.4, 0.85)` amplified the square
edges into hard-edged glowing cubes.
* Fix: shared 128×128 radial-gradient `CanvasTexture` singleton used
as the sprite map. Retuned bloom to `(0.55, 0.6, 0.2)`. Halved fog
density (0.008 → 0.0035). Edges bumped from dark navy `0x4a4a7a`
to brand violet `0x8b5cf6` with higher opacity. Added explicit
`scene.background` and a 2000-point starfield for depth.
* 21 regression tests added in `ui-fixes.test.ts` locking every
invariant in (shared texture singleton, depthWrite:false, scale
×6, bloom magic numbers via source regex, starfield presence).
Tests: 1,284 Rust (+47) + 171 Vitest (+21) = 1,455 total, 0 failed
Clippy: clean across all targets, zero warnings
Release binary: 22.6MB, `cargo build --release -p vestige-mcp` green
Versions: workspace aligned at 2.0.5 across all 6 crates/packages
Closes #31
2026-04-14 17:30:30 -05:00
let before = storage
. get_node ( & args . id )
. map_err ( | e | e . to_string ( ) ) ?
2026-01-25 13:30:03 -06:00
. ok_or_else ( | | format! ( " Node not found: {} " , args . id ) ) ? ;
feat(v2.0.5): Intentional Amnesia — active forgetting via top-down inhibitory control
First AI memory system to model forgetting as a neuroscience-grounded
PROCESS rather than passive decay. Adds the `suppress` MCP tool (#24),
Rac1 cascade worker, migration V10, and dashboard forgetting indicators.
Based on:
- Anderson, Hanslmayr & Quaegebeur (2025), Nat Rev Neurosci — right
lateral PFC as the domain-general inhibitory controller; SIF
compounds with each stopping attempt.
- Cervantes-Sandoval et al. (2020), Front Cell Neurosci PMC7477079 —
Rac1 GTPase as the active synaptic destabilization mechanism.
What's new:
* `suppress` MCP tool — each call compounds `suppression_count` and
subtracts a `0.15 × count` penalty (saturating at 80%) from
retrieval scores during hybrid search. Distinct from delete
(removes) and demote (one-shot).
* Rac1 cascade worker — background sweep piggybacks the 6h
consolidation loop, walks `memory_connections` edges from
recently-suppressed seeds, applies attenuated FSRS decay to
co-activated neighbors. You don't just forget Jake — you fade
the café, the roommate, the birthday.
* 24h labile window — reversible via `suppress({id, reverse: true})`
within 24 hours. Matches Nader reconsolidation semantics.
* Migration V10 — additive-only (`suppression_count`, `suppressed_at`
+ partial indices). All v2.0.x DBs upgrade seamlessly on first launch.
* Dashboard: `ForgettingIndicator.svelte` pulses when suppressions
are active. 3D graph nodes dim to 20% opacity when suppressed.
New WebSocket events: `MemorySuppressed`, `MemoryUnsuppressed`,
`Rac1CascadeSwept`. Heartbeat carries `suppressed_count`.
* Search pipeline: SIF penalty inserted into the accessibility stage
so it stacks on top of passive FSRS decay.
* Tool count bumped 23 → 24. Cognitive modules 29 → 30.
Memories persist — they are INHIBITED, not erased. `memory.get(id)`
returns full content through any number of suppressions. The 24h
labile window is a grace period for regret.
Also fixes issue #31 (dashboard graph view buggy) as a companion UI
bug discovered during the v2.0.5 audit cycle:
* Root cause: node glow `SpriteMaterial` had no `map`, so
`THREE.Sprite` rendered as a solid-coloured 1×1 plane. Additive
blending + `UnrealBloomPass(0.8, 0.4, 0.85)` amplified the square
edges into hard-edged glowing cubes.
* Fix: shared 128×128 radial-gradient `CanvasTexture` singleton used
as the sprite map. Retuned bloom to `(0.55, 0.6, 0.2)`. Halved fog
density (0.008 → 0.0035). Edges bumped from dark navy `0x4a4a7a`
to brand violet `0x8b5cf6` with higher opacity. Added explicit
`scene.background` and a 2000-point starfield for depth.
* 21 regression tests added in `ui-fixes.test.ts` locking every
invariant in (shared texture singleton, depthWrite:false, scale
×6, bloom magic numbers via source regex, starfield presence).
Tests: 1,284 Rust (+47) + 171 Vitest (+21) = 1,455 total, 0 failed
Clippy: clean across all targets, zero warnings
Release binary: 22.6MB, `cargo build --release -p vestige-mcp` green
Versions: workspace aligned at 2.0.5 across all 6 crates/packages
Closes #31
2026-04-14 17:30:30 -05:00
let node = storage
. promote_memory ( & args . id )
. map_err ( | e | e . to_string ( ) ) ? ;
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
// ====================================================================
// COGNITIVE FEEDBACK PIPELINE (promote)
// ====================================================================
if let Ok ( mut cog ) = cognitive . try_lock ( ) {
// 5A. Reward signal — record positive outcome
feat(v2.0.5): Intentional Amnesia — active forgetting via top-down inhibitory control
First AI memory system to model forgetting as a neuroscience-grounded
PROCESS rather than passive decay. Adds the `suppress` MCP tool (#24),
Rac1 cascade worker, migration V10, and dashboard forgetting indicators.
Based on:
- Anderson, Hanslmayr & Quaegebeur (2025), Nat Rev Neurosci — right
lateral PFC as the domain-general inhibitory controller; SIF
compounds with each stopping attempt.
- Cervantes-Sandoval et al. (2020), Front Cell Neurosci PMC7477079 —
Rac1 GTPase as the active synaptic destabilization mechanism.
What's new:
* `suppress` MCP tool — each call compounds `suppression_count` and
subtracts a `0.15 × count` penalty (saturating at 80%) from
retrieval scores during hybrid search. Distinct from delete
(removes) and demote (one-shot).
* Rac1 cascade worker — background sweep piggybacks the 6h
consolidation loop, walks `memory_connections` edges from
recently-suppressed seeds, applies attenuated FSRS decay to
co-activated neighbors. You don't just forget Jake — you fade
the café, the roommate, the birthday.
* 24h labile window — reversible via `suppress({id, reverse: true})`
within 24 hours. Matches Nader reconsolidation semantics.
* Migration V10 — additive-only (`suppression_count`, `suppressed_at`
+ partial indices). All v2.0.x DBs upgrade seamlessly on first launch.
* Dashboard: `ForgettingIndicator.svelte` pulses when suppressions
are active. 3D graph nodes dim to 20% opacity when suppressed.
New WebSocket events: `MemorySuppressed`, `MemoryUnsuppressed`,
`Rac1CascadeSwept`. Heartbeat carries `suppressed_count`.
* Search pipeline: SIF penalty inserted into the accessibility stage
so it stacks on top of passive FSRS decay.
* Tool count bumped 23 → 24. Cognitive modules 29 → 30.
Memories persist — they are INHIBITED, not erased. `memory.get(id)`
returns full content through any number of suppressions. The 24h
labile window is a grace period for regret.
Also fixes issue #31 (dashboard graph view buggy) as a companion UI
bug discovered during the v2.0.5 audit cycle:
* Root cause: node glow `SpriteMaterial` had no `map`, so
`THREE.Sprite` rendered as a solid-coloured 1×1 plane. Additive
blending + `UnrealBloomPass(0.8, 0.4, 0.85)` amplified the square
edges into hard-edged glowing cubes.
* Fix: shared 128×128 radial-gradient `CanvasTexture` singleton used
as the sprite map. Retuned bloom to `(0.55, 0.6, 0.2)`. Halved fog
density (0.008 → 0.0035). Edges bumped from dark navy `0x4a4a7a`
to brand violet `0x8b5cf6` with higher opacity. Added explicit
`scene.background` and a 2000-point starfield for depth.
* 21 regression tests added in `ui-fixes.test.ts` locking every
invariant in (shared texture singleton, depthWrite:false, scale
×6, bloom magic numbers via source regex, starfield presence).
Tests: 1,284 Rust (+47) + 171 Vitest (+21) = 1,455 total, 0 failed
Clippy: clean across all targets, zero warnings
Release binary: 22.6MB, `cargo build --release -p vestige-mcp` green
Versions: workspace aligned at 2.0.5 across all 6 crates/packages
Closes #31
2026-04-14 17:30:30 -05:00
cog . reward_signal
. record_outcome ( & args . id , OutcomeType ::Helpful ) ;
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
// 5B. Importance tracking — mark as helpful retrieval
cog . importance_tracker . on_retrieved ( & args . id , true ) ;
// 5C. Reconsolidation — boost retrieval if memory is labile
if cog . reconsolidation . is_labile ( & args . id ) {
cog . reconsolidation . apply_modification (
& args . id ,
Modification ::StrengthenConnection {
target_memory_id : args . id . clone ( ) ,
boost : 0.2 ,
} ,
) ;
}
}
2026-01-25 13:30:03 -06:00
Ok ( serde_json ::json! ( {
" success " : true ,
" action " : " promoted " ,
" nodeId " : node . id ,
" reason " : args . reason ,
" changes " : {
" retrievalStrength " : {
" before " : before . retrieval_strength ,
" after " : node . retrieval_strength ,
" delta " : " +0.20 "
} ,
" retentionStrength " : {
" before " : before . retention_strength ,
" after " : node . retention_strength ,
" delta " : " +0.10 "
} ,
" stability " : {
" before " : before . stability ,
" after " : node . stability ,
" multiplier " : " 1.5x "
}
} ,
" message " : format ! ( " Memory promoted. It will now surface more often in searches. Retrieval: {:.2} -> {:.2} " ,
before . retrieval_strength , node . retrieval_strength ) ,
} ) )
}
/// Demote a memory (thumbs down) - it led to a bad outcome
pub async fn execute_demote (
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
storage : & Arc < Storage > ,
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
cognitive : & Arc < Mutex < CognitiveEngine > > ,
2026-01-25 13:30:03 -06:00
args : Option < Value > ,
) -> Result < Value , String > {
let args : FeedbackArgs = match args {
Some ( v ) = > serde_json ::from_value ( v ) . map_err ( | e | format! ( " Invalid arguments: {} " , e ) ) ? ,
None = > return Err ( " Missing arguments " . to_string ( ) ) ,
} ;
// Validate UUID
uuid ::Uuid ::parse_str ( & args . id ) . map_err ( | _ | " Invalid node ID format " . to_string ( ) ) ? ;
// Get node before for comparison
feat(v2.0.5): Intentional Amnesia — active forgetting via top-down inhibitory control
First AI memory system to model forgetting as a neuroscience-grounded
PROCESS rather than passive decay. Adds the `suppress` MCP tool (#24),
Rac1 cascade worker, migration V10, and dashboard forgetting indicators.
Based on:
- Anderson, Hanslmayr & Quaegebeur (2025), Nat Rev Neurosci — right
lateral PFC as the domain-general inhibitory controller; SIF
compounds with each stopping attempt.
- Cervantes-Sandoval et al. (2020), Front Cell Neurosci PMC7477079 —
Rac1 GTPase as the active synaptic destabilization mechanism.
What's new:
* `suppress` MCP tool — each call compounds `suppression_count` and
subtracts a `0.15 × count` penalty (saturating at 80%) from
retrieval scores during hybrid search. Distinct from delete
(removes) and demote (one-shot).
* Rac1 cascade worker — background sweep piggybacks the 6h
consolidation loop, walks `memory_connections` edges from
recently-suppressed seeds, applies attenuated FSRS decay to
co-activated neighbors. You don't just forget Jake — you fade
the café, the roommate, the birthday.
* 24h labile window — reversible via `suppress({id, reverse: true})`
within 24 hours. Matches Nader reconsolidation semantics.
* Migration V10 — additive-only (`suppression_count`, `suppressed_at`
+ partial indices). All v2.0.x DBs upgrade seamlessly on first launch.
* Dashboard: `ForgettingIndicator.svelte` pulses when suppressions
are active. 3D graph nodes dim to 20% opacity when suppressed.
New WebSocket events: `MemorySuppressed`, `MemoryUnsuppressed`,
`Rac1CascadeSwept`. Heartbeat carries `suppressed_count`.
* Search pipeline: SIF penalty inserted into the accessibility stage
so it stacks on top of passive FSRS decay.
* Tool count bumped 23 → 24. Cognitive modules 29 → 30.
Memories persist — they are INHIBITED, not erased. `memory.get(id)`
returns full content through any number of suppressions. The 24h
labile window is a grace period for regret.
Also fixes issue #31 (dashboard graph view buggy) as a companion UI
bug discovered during the v2.0.5 audit cycle:
* Root cause: node glow `SpriteMaterial` had no `map`, so
`THREE.Sprite` rendered as a solid-coloured 1×1 plane. Additive
blending + `UnrealBloomPass(0.8, 0.4, 0.85)` amplified the square
edges into hard-edged glowing cubes.
* Fix: shared 128×128 radial-gradient `CanvasTexture` singleton used
as the sprite map. Retuned bloom to `(0.55, 0.6, 0.2)`. Halved fog
density (0.008 → 0.0035). Edges bumped from dark navy `0x4a4a7a`
to brand violet `0x8b5cf6` with higher opacity. Added explicit
`scene.background` and a 2000-point starfield for depth.
* 21 regression tests added in `ui-fixes.test.ts` locking every
invariant in (shared texture singleton, depthWrite:false, scale
×6, bloom magic numbers via source regex, starfield presence).
Tests: 1,284 Rust (+47) + 171 Vitest (+21) = 1,455 total, 0 failed
Clippy: clean across all targets, zero warnings
Release binary: 22.6MB, `cargo build --release -p vestige-mcp` green
Versions: workspace aligned at 2.0.5 across all 6 crates/packages
Closes #31
2026-04-14 17:30:30 -05:00
let before = storage
. get_node ( & args . id )
. map_err ( | e | e . to_string ( ) ) ?
2026-01-25 13:30:03 -06:00
. ok_or_else ( | | format! ( " Node not found: {} " , args . id ) ) ? ;
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
let node = storage . demote_memory ( & args . id ) . map_err ( | e | e . to_string ( ) ) ? ;
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
// ====================================================================
// COGNITIVE FEEDBACK PIPELINE (demote)
// ====================================================================
if let Ok ( mut cog ) = cognitive . try_lock ( ) {
// 5A. Reward signal — record negative outcome
feat(v2.0.5): Intentional Amnesia — active forgetting via top-down inhibitory control
First AI memory system to model forgetting as a neuroscience-grounded
PROCESS rather than passive decay. Adds the `suppress` MCP tool (#24),
Rac1 cascade worker, migration V10, and dashboard forgetting indicators.
Based on:
- Anderson, Hanslmayr & Quaegebeur (2025), Nat Rev Neurosci — right
lateral PFC as the domain-general inhibitory controller; SIF
compounds with each stopping attempt.
- Cervantes-Sandoval et al. (2020), Front Cell Neurosci PMC7477079 —
Rac1 GTPase as the active synaptic destabilization mechanism.
What's new:
* `suppress` MCP tool — each call compounds `suppression_count` and
subtracts a `0.15 × count` penalty (saturating at 80%) from
retrieval scores during hybrid search. Distinct from delete
(removes) and demote (one-shot).
* Rac1 cascade worker — background sweep piggybacks the 6h
consolidation loop, walks `memory_connections` edges from
recently-suppressed seeds, applies attenuated FSRS decay to
co-activated neighbors. You don't just forget Jake — you fade
the café, the roommate, the birthday.
* 24h labile window — reversible via `suppress({id, reverse: true})`
within 24 hours. Matches Nader reconsolidation semantics.
* Migration V10 — additive-only (`suppression_count`, `suppressed_at`
+ partial indices). All v2.0.x DBs upgrade seamlessly on first launch.
* Dashboard: `ForgettingIndicator.svelte` pulses when suppressions
are active. 3D graph nodes dim to 20% opacity when suppressed.
New WebSocket events: `MemorySuppressed`, `MemoryUnsuppressed`,
`Rac1CascadeSwept`. Heartbeat carries `suppressed_count`.
* Search pipeline: SIF penalty inserted into the accessibility stage
so it stacks on top of passive FSRS decay.
* Tool count bumped 23 → 24. Cognitive modules 29 → 30.
Memories persist — they are INHIBITED, not erased. `memory.get(id)`
returns full content through any number of suppressions. The 24h
labile window is a grace period for regret.
Also fixes issue #31 (dashboard graph view buggy) as a companion UI
bug discovered during the v2.0.5 audit cycle:
* Root cause: node glow `SpriteMaterial` had no `map`, so
`THREE.Sprite` rendered as a solid-coloured 1×1 plane. Additive
blending + `UnrealBloomPass(0.8, 0.4, 0.85)` amplified the square
edges into hard-edged glowing cubes.
* Fix: shared 128×128 radial-gradient `CanvasTexture` singleton used
as the sprite map. Retuned bloom to `(0.55, 0.6, 0.2)`. Halved fog
density (0.008 → 0.0035). Edges bumped from dark navy `0x4a4a7a`
to brand violet `0x8b5cf6` with higher opacity. Added explicit
`scene.background` and a 2000-point starfield for depth.
* 21 regression tests added in `ui-fixes.test.ts` locking every
invariant in (shared texture singleton, depthWrite:false, scale
×6, bloom magic numbers via source regex, starfield presence).
Tests: 1,284 Rust (+47) + 171 Vitest (+21) = 1,455 total, 0 failed
Clippy: clean across all targets, zero warnings
Release binary: 22.6MB, `cargo build --release -p vestige-mcp` green
Versions: workspace aligned at 2.0.5 across all 6 crates/packages
Closes #31
2026-04-14 17:30:30 -05:00
cog . reward_signal
. record_outcome ( & args . id , OutcomeType ::NotHelpful ) ;
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
// 5B. Importance tracking — mark as unhelpful retrieval
cog . importance_tracker . on_retrieved ( & args . id , false ) ;
// 5C. Reconsolidation — weaken if memory is labile
if cog . reconsolidation . is_labile ( & args . id ) {
cog . reconsolidation . apply_modification (
& args . id ,
Modification ::AddContext {
context : " User reported this memory was wrong/unhelpful " . to_string ( ) ,
} ,
) ;
}
}
2026-01-25 13:30:03 -06:00
Ok ( serde_json ::json! ( {
" success " : true ,
" action " : " demoted " ,
" nodeId " : node . id ,
" reason " : args . reason ,
" changes " : {
" retrievalStrength " : {
" before " : before . retrieval_strength ,
" after " : node . retrieval_strength ,
" delta " : " -0.30 "
} ,
" retentionStrength " : {
" before " : before . retention_strength ,
" after " : node . retention_strength ,
" delta " : " -0.15 "
} ,
" stability " : {
" before " : before . stability ,
" after " : node . stability ,
" multiplier " : " 0.5x "
}
} ,
" message " : format ! ( " Memory demoted. Better alternatives will now surface instead. Retrieval: {:.2} -> {:.2} " ,
before . retrieval_strength , node . retrieval_strength ) ,
" note " : " Memory is NOT deleted - it remains searchable but ranks lower. "
} ) )
}
/// Input schema for request_feedback tool
pub fn request_feedback_schema ( ) -> Value {
serde_json ::json! ( {
" type " : " object " ,
" properties " : {
" id " : {
" type " : " string " ,
" description " : " The ID of the memory to request feedback on "
} ,
" context " : {
" type " : " string " ,
" description " : " What the memory was used for (e.g., 'error handling advice') "
}
} ,
" required " : [ " id " ]
} )
}
#[ derive(Debug, Deserialize) ]
struct RequestFeedbackArgs {
id : String ,
context : Option < String > ,
}
/// Request feedback from the user about a memory's usefulness
/// Returns a structured prompt for Claude to ask the user
pub async fn execute_request_feedback (
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
storage : & Arc < Storage > ,
2026-01-25 13:30:03 -06:00
args : Option < Value > ,
) -> Result < Value , String > {
let args : RequestFeedbackArgs = match args {
Some ( v ) = > serde_json ::from_value ( v ) . map_err ( | e | format! ( " Invalid arguments: {} " , e ) ) ? ,
None = > return Err ( " Missing arguments " . to_string ( ) ) ,
} ;
// Validate UUID
uuid ::Uuid ::parse_str ( & args . id ) . map_err ( | _ | " Invalid node ID format " . to_string ( ) ) ? ;
feat(v2.0.5): Intentional Amnesia — active forgetting via top-down inhibitory control
First AI memory system to model forgetting as a neuroscience-grounded
PROCESS rather than passive decay. Adds the `suppress` MCP tool (#24),
Rac1 cascade worker, migration V10, and dashboard forgetting indicators.
Based on:
- Anderson, Hanslmayr & Quaegebeur (2025), Nat Rev Neurosci — right
lateral PFC as the domain-general inhibitory controller; SIF
compounds with each stopping attempt.
- Cervantes-Sandoval et al. (2020), Front Cell Neurosci PMC7477079 —
Rac1 GTPase as the active synaptic destabilization mechanism.
What's new:
* `suppress` MCP tool — each call compounds `suppression_count` and
subtracts a `0.15 × count` penalty (saturating at 80%) from
retrieval scores during hybrid search. Distinct from delete
(removes) and demote (one-shot).
* Rac1 cascade worker — background sweep piggybacks the 6h
consolidation loop, walks `memory_connections` edges from
recently-suppressed seeds, applies attenuated FSRS decay to
co-activated neighbors. You don't just forget Jake — you fade
the café, the roommate, the birthday.
* 24h labile window — reversible via `suppress({id, reverse: true})`
within 24 hours. Matches Nader reconsolidation semantics.
* Migration V10 — additive-only (`suppression_count`, `suppressed_at`
+ partial indices). All v2.0.x DBs upgrade seamlessly on first launch.
* Dashboard: `ForgettingIndicator.svelte` pulses when suppressions
are active. 3D graph nodes dim to 20% opacity when suppressed.
New WebSocket events: `MemorySuppressed`, `MemoryUnsuppressed`,
`Rac1CascadeSwept`. Heartbeat carries `suppressed_count`.
* Search pipeline: SIF penalty inserted into the accessibility stage
so it stacks on top of passive FSRS decay.
* Tool count bumped 23 → 24. Cognitive modules 29 → 30.
Memories persist — they are INHIBITED, not erased. `memory.get(id)`
returns full content through any number of suppressions. The 24h
labile window is a grace period for regret.
Also fixes issue #31 (dashboard graph view buggy) as a companion UI
bug discovered during the v2.0.5 audit cycle:
* Root cause: node glow `SpriteMaterial` had no `map`, so
`THREE.Sprite` rendered as a solid-coloured 1×1 plane. Additive
blending + `UnrealBloomPass(0.8, 0.4, 0.85)` amplified the square
edges into hard-edged glowing cubes.
* Fix: shared 128×128 radial-gradient `CanvasTexture` singleton used
as the sprite map. Retuned bloom to `(0.55, 0.6, 0.2)`. Halved fog
density (0.008 → 0.0035). Edges bumped from dark navy `0x4a4a7a`
to brand violet `0x8b5cf6` with higher opacity. Added explicit
`scene.background` and a 2000-point starfield for depth.
* 21 regression tests added in `ui-fixes.test.ts` locking every
invariant in (shared texture singleton, depthWrite:false, scale
×6, bloom magic numbers via source regex, starfield presence).
Tests: 1,284 Rust (+47) + 171 Vitest (+21) = 1,455 total, 0 failed
Clippy: clean across all targets, zero warnings
Release binary: 22.6MB, `cargo build --release -p vestige-mcp` green
Versions: workspace aligned at 2.0.5 across all 6 crates/packages
Closes #31
2026-04-14 17:30:30 -05:00
let node = storage
. get_node ( & args . id )
. map_err ( | e | e . to_string ( ) ) ?
2026-01-25 13:30:03 -06:00
. ok_or_else ( | | format! ( " Node not found: {} " , args . id ) ) ? ;
// Truncate content for display
let preview : String = node . content . chars ( ) . take ( 100 ) . collect ( ) ;
let preview = if node . content . len ( ) > 100 {
format! ( " {} ... " , preview )
} else {
preview
} ;
Ok ( serde_json ::json! ( {
" action " : " request_feedback " ,
" nodeId " : node . id ,
" memoryPreview " : preview ,
" context " : args . context ,
" prompt " : " Was this memory helpful? " ,
" options " : [
{
" key " : " A " ,
" label " : " Yes, helpful " ,
" action " : " promote " ,
" description " : " Memory will surface more often "
} ,
{
" key " : " B " ,
" label " : " No, wrong/outdated " ,
" action " : " demote " ,
" description " : " Better alternatives will surface instead "
} ,
{
" key " : " C " ,
" label " : " Ask Claude... " ,
" action " : " custom " ,
" description " : " Give Claude a custom instruction (e.g., 'update this memory', 'merge with X', 'add tag Y') "
}
] ,
2026-02-20 21:59:52 -06:00
" instruction " : " PRESENT THESE OPTIONS TO THE USER. If they choose A, call memory(action='promote'). If B, call memory(action='demote'). If C, they will provide a custom instruction - execute it (could be: update the memory content, delete it, merge it, add tags, research something, etc.). "
2026-01-25 13:30:03 -06:00
} ) )
}
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
#[ cfg(test) ]
mod tests {
use super ::* ;
use crate ::cognitive ::CognitiveEngine ;
use tempfile ::TempDir ;
fn test_cognitive ( ) -> Arc < Mutex < CognitiveEngine > > {
Arc ::new ( Mutex ::new ( CognitiveEngine ::new ( ) ) )
}
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
async fn test_storage ( ) -> ( Arc < Storage > , TempDir ) {
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
let dir = TempDir ::new ( ) . unwrap ( ) ;
let storage = Storage ::new ( Some ( dir . path ( ) . join ( " test.db " ) ) ) . unwrap ( ) ;
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
( Arc ::new ( storage ) , dir )
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
}
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
async fn ingest_test_memory ( storage : & Arc < Storage > ) -> String {
let node = storage
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
. ingest ( vestige_core ::IngestInput {
content : " Test memory for feedback " . to_string ( ) ,
node_type : " fact " . to_string ( ) ,
source : None ,
sentiment_score : 0.0 ,
sentiment_magnitude : 0.0 ,
tags : vec ! [ ] ,
valid_from : None ,
valid_until : None ,
} )
. unwrap ( ) ;
node . id
}
// === PROMOTE SCHEMA ===
#[ test ]
fn test_promote_schema_has_required_fields ( ) {
let schema = promote_schema ( ) ;
assert_eq! ( schema [ " type " ] , " object " ) ;
assert! ( schema [ " properties " ] [ " id " ] . is_object ( ) ) ;
assert! ( schema [ " properties " ] [ " reason " ] . is_object ( ) ) ;
feat(v2.0.5): Intentional Amnesia — active forgetting via top-down inhibitory control
First AI memory system to model forgetting as a neuroscience-grounded
PROCESS rather than passive decay. Adds the `suppress` MCP tool (#24),
Rac1 cascade worker, migration V10, and dashboard forgetting indicators.
Based on:
- Anderson, Hanslmayr & Quaegebeur (2025), Nat Rev Neurosci — right
lateral PFC as the domain-general inhibitory controller; SIF
compounds with each stopping attempt.
- Cervantes-Sandoval et al. (2020), Front Cell Neurosci PMC7477079 —
Rac1 GTPase as the active synaptic destabilization mechanism.
What's new:
* `suppress` MCP tool — each call compounds `suppression_count` and
subtracts a `0.15 × count` penalty (saturating at 80%) from
retrieval scores during hybrid search. Distinct from delete
(removes) and demote (one-shot).
* Rac1 cascade worker — background sweep piggybacks the 6h
consolidation loop, walks `memory_connections` edges from
recently-suppressed seeds, applies attenuated FSRS decay to
co-activated neighbors. You don't just forget Jake — you fade
the café, the roommate, the birthday.
* 24h labile window — reversible via `suppress({id, reverse: true})`
within 24 hours. Matches Nader reconsolidation semantics.
* Migration V10 — additive-only (`suppression_count`, `suppressed_at`
+ partial indices). All v2.0.x DBs upgrade seamlessly on first launch.
* Dashboard: `ForgettingIndicator.svelte` pulses when suppressions
are active. 3D graph nodes dim to 20% opacity when suppressed.
New WebSocket events: `MemorySuppressed`, `MemoryUnsuppressed`,
`Rac1CascadeSwept`. Heartbeat carries `suppressed_count`.
* Search pipeline: SIF penalty inserted into the accessibility stage
so it stacks on top of passive FSRS decay.
* Tool count bumped 23 → 24. Cognitive modules 29 → 30.
Memories persist — they are INHIBITED, not erased. `memory.get(id)`
returns full content through any number of suppressions. The 24h
labile window is a grace period for regret.
Also fixes issue #31 (dashboard graph view buggy) as a companion UI
bug discovered during the v2.0.5 audit cycle:
* Root cause: node glow `SpriteMaterial` had no `map`, so
`THREE.Sprite` rendered as a solid-coloured 1×1 plane. Additive
blending + `UnrealBloomPass(0.8, 0.4, 0.85)` amplified the square
edges into hard-edged glowing cubes.
* Fix: shared 128×128 radial-gradient `CanvasTexture` singleton used
as the sprite map. Retuned bloom to `(0.55, 0.6, 0.2)`. Halved fog
density (0.008 → 0.0035). Edges bumped from dark navy `0x4a4a7a`
to brand violet `0x8b5cf6` with higher opacity. Added explicit
`scene.background` and a 2000-point starfield for depth.
* 21 regression tests added in `ui-fixes.test.ts` locking every
invariant in (shared texture singleton, depthWrite:false, scale
×6, bloom magic numbers via source regex, starfield presence).
Tests: 1,284 Rust (+47) + 171 Vitest (+21) = 1,455 total, 0 failed
Clippy: clean across all targets, zero warnings
Release binary: 22.6MB, `cargo build --release -p vestige-mcp` green
Versions: workspace aligned at 2.0.5 across all 6 crates/packages
Closes #31
2026-04-14 17:30:30 -05:00
assert! (
schema [ " required " ]
. as_array ( )
. unwrap ( )
. contains ( & serde_json ::json! ( " id " ) )
) ;
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
}
#[ test ]
fn test_demote_schema_has_required_fields ( ) {
let schema = demote_schema ( ) ;
assert_eq! ( schema [ " type " ] , " object " ) ;
assert! ( schema [ " properties " ] [ " id " ] . is_object ( ) ) ;
feat(v2.0.5): Intentional Amnesia — active forgetting via top-down inhibitory control
First AI memory system to model forgetting as a neuroscience-grounded
PROCESS rather than passive decay. Adds the `suppress` MCP tool (#24),
Rac1 cascade worker, migration V10, and dashboard forgetting indicators.
Based on:
- Anderson, Hanslmayr & Quaegebeur (2025), Nat Rev Neurosci — right
lateral PFC as the domain-general inhibitory controller; SIF
compounds with each stopping attempt.
- Cervantes-Sandoval et al. (2020), Front Cell Neurosci PMC7477079 —
Rac1 GTPase as the active synaptic destabilization mechanism.
What's new:
* `suppress` MCP tool — each call compounds `suppression_count` and
subtracts a `0.15 × count` penalty (saturating at 80%) from
retrieval scores during hybrid search. Distinct from delete
(removes) and demote (one-shot).
* Rac1 cascade worker — background sweep piggybacks the 6h
consolidation loop, walks `memory_connections` edges from
recently-suppressed seeds, applies attenuated FSRS decay to
co-activated neighbors. You don't just forget Jake — you fade
the café, the roommate, the birthday.
* 24h labile window — reversible via `suppress({id, reverse: true})`
within 24 hours. Matches Nader reconsolidation semantics.
* Migration V10 — additive-only (`suppression_count`, `suppressed_at`
+ partial indices). All v2.0.x DBs upgrade seamlessly on first launch.
* Dashboard: `ForgettingIndicator.svelte` pulses when suppressions
are active. 3D graph nodes dim to 20% opacity when suppressed.
New WebSocket events: `MemorySuppressed`, `MemoryUnsuppressed`,
`Rac1CascadeSwept`. Heartbeat carries `suppressed_count`.
* Search pipeline: SIF penalty inserted into the accessibility stage
so it stacks on top of passive FSRS decay.
* Tool count bumped 23 → 24. Cognitive modules 29 → 30.
Memories persist — they are INHIBITED, not erased. `memory.get(id)`
returns full content through any number of suppressions. The 24h
labile window is a grace period for regret.
Also fixes issue #31 (dashboard graph view buggy) as a companion UI
bug discovered during the v2.0.5 audit cycle:
* Root cause: node glow `SpriteMaterial` had no `map`, so
`THREE.Sprite` rendered as a solid-coloured 1×1 plane. Additive
blending + `UnrealBloomPass(0.8, 0.4, 0.85)` amplified the square
edges into hard-edged glowing cubes.
* Fix: shared 128×128 radial-gradient `CanvasTexture` singleton used
as the sprite map. Retuned bloom to `(0.55, 0.6, 0.2)`. Halved fog
density (0.008 → 0.0035). Edges bumped from dark navy `0x4a4a7a`
to brand violet `0x8b5cf6` with higher opacity. Added explicit
`scene.background` and a 2000-point starfield for depth.
* 21 regression tests added in `ui-fixes.test.ts` locking every
invariant in (shared texture singleton, depthWrite:false, scale
×6, bloom magic numbers via source regex, starfield presence).
Tests: 1,284 Rust (+47) + 171 Vitest (+21) = 1,455 total, 0 failed
Clippy: clean across all targets, zero warnings
Release binary: 22.6MB, `cargo build --release -p vestige-mcp` green
Versions: workspace aligned at 2.0.5 across all 6 crates/packages
Closes #31
2026-04-14 17:30:30 -05:00
assert! (
schema [ " required " ]
. as_array ( )
. unwrap ( )
. contains ( & serde_json ::json! ( " id " ) )
) ;
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
}
#[ test ]
fn test_request_feedback_schema_has_required_fields ( ) {
let schema = request_feedback_schema ( ) ;
assert_eq! ( schema [ " type " ] , " object " ) ;
assert! ( schema [ " properties " ] [ " id " ] . is_object ( ) ) ;
assert! ( schema [ " properties " ] [ " context " ] . is_object ( ) ) ;
feat(v2.0.5): Intentional Amnesia — active forgetting via top-down inhibitory control
First AI memory system to model forgetting as a neuroscience-grounded
PROCESS rather than passive decay. Adds the `suppress` MCP tool (#24),
Rac1 cascade worker, migration V10, and dashboard forgetting indicators.
Based on:
- Anderson, Hanslmayr & Quaegebeur (2025), Nat Rev Neurosci — right
lateral PFC as the domain-general inhibitory controller; SIF
compounds with each stopping attempt.
- Cervantes-Sandoval et al. (2020), Front Cell Neurosci PMC7477079 —
Rac1 GTPase as the active synaptic destabilization mechanism.
What's new:
* `suppress` MCP tool — each call compounds `suppression_count` and
subtracts a `0.15 × count` penalty (saturating at 80%) from
retrieval scores during hybrid search. Distinct from delete
(removes) and demote (one-shot).
* Rac1 cascade worker — background sweep piggybacks the 6h
consolidation loop, walks `memory_connections` edges from
recently-suppressed seeds, applies attenuated FSRS decay to
co-activated neighbors. You don't just forget Jake — you fade
the café, the roommate, the birthday.
* 24h labile window — reversible via `suppress({id, reverse: true})`
within 24 hours. Matches Nader reconsolidation semantics.
* Migration V10 — additive-only (`suppression_count`, `suppressed_at`
+ partial indices). All v2.0.x DBs upgrade seamlessly on first launch.
* Dashboard: `ForgettingIndicator.svelte` pulses when suppressions
are active. 3D graph nodes dim to 20% opacity when suppressed.
New WebSocket events: `MemorySuppressed`, `MemoryUnsuppressed`,
`Rac1CascadeSwept`. Heartbeat carries `suppressed_count`.
* Search pipeline: SIF penalty inserted into the accessibility stage
so it stacks on top of passive FSRS decay.
* Tool count bumped 23 → 24. Cognitive modules 29 → 30.
Memories persist — they are INHIBITED, not erased. `memory.get(id)`
returns full content through any number of suppressions. The 24h
labile window is a grace period for regret.
Also fixes issue #31 (dashboard graph view buggy) as a companion UI
bug discovered during the v2.0.5 audit cycle:
* Root cause: node glow `SpriteMaterial` had no `map`, so
`THREE.Sprite` rendered as a solid-coloured 1×1 plane. Additive
blending + `UnrealBloomPass(0.8, 0.4, 0.85)` amplified the square
edges into hard-edged glowing cubes.
* Fix: shared 128×128 radial-gradient `CanvasTexture` singleton used
as the sprite map. Retuned bloom to `(0.55, 0.6, 0.2)`. Halved fog
density (0.008 → 0.0035). Edges bumped from dark navy `0x4a4a7a`
to brand violet `0x8b5cf6` with higher opacity. Added explicit
`scene.background` and a 2000-point starfield for depth.
* 21 regression tests added in `ui-fixes.test.ts` locking every
invariant in (shared texture singleton, depthWrite:false, scale
×6, bloom magic numbers via source regex, starfield presence).
Tests: 1,284 Rust (+47) + 171 Vitest (+21) = 1,455 total, 0 failed
Clippy: clean across all targets, zero warnings
Release binary: 22.6MB, `cargo build --release -p vestige-mcp` green
Versions: workspace aligned at 2.0.5 across all 6 crates/packages
Closes #31
2026-04-14 17:30:30 -05:00
assert! (
schema [ " required " ]
. as_array ( )
. unwrap ( )
. contains ( & serde_json ::json! ( " id " ) )
) ;
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
}
// === PROMOTE TESTS ===
#[ tokio::test ]
async fn test_promote_missing_args_fails ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
let result = execute_promote ( & storage , & test_cognitive ( ) , None ) . await ;
assert! ( result . is_err ( ) ) ;
assert! ( result . unwrap_err ( ) . contains ( " Missing arguments " ) ) ;
}
#[ tokio::test ]
async fn test_promote_invalid_uuid_fails ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
let args = serde_json ::json! ( { " id " : " not-a-uuid " } ) ;
let result = execute_promote ( & storage , & test_cognitive ( ) , Some ( args ) ) . await ;
assert! ( result . is_err ( ) ) ;
assert! ( result . unwrap_err ( ) . contains ( " Invalid node ID format " ) ) ;
}
#[ tokio::test ]
async fn test_promote_nonexistent_node_fails ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
feat(v2.0.5): Intentional Amnesia — active forgetting via top-down inhibitory control
First AI memory system to model forgetting as a neuroscience-grounded
PROCESS rather than passive decay. Adds the `suppress` MCP tool (#24),
Rac1 cascade worker, migration V10, and dashboard forgetting indicators.
Based on:
- Anderson, Hanslmayr & Quaegebeur (2025), Nat Rev Neurosci — right
lateral PFC as the domain-general inhibitory controller; SIF
compounds with each stopping attempt.
- Cervantes-Sandoval et al. (2020), Front Cell Neurosci PMC7477079 —
Rac1 GTPase as the active synaptic destabilization mechanism.
What's new:
* `suppress` MCP tool — each call compounds `suppression_count` and
subtracts a `0.15 × count` penalty (saturating at 80%) from
retrieval scores during hybrid search. Distinct from delete
(removes) and demote (one-shot).
* Rac1 cascade worker — background sweep piggybacks the 6h
consolidation loop, walks `memory_connections` edges from
recently-suppressed seeds, applies attenuated FSRS decay to
co-activated neighbors. You don't just forget Jake — you fade
the café, the roommate, the birthday.
* 24h labile window — reversible via `suppress({id, reverse: true})`
within 24 hours. Matches Nader reconsolidation semantics.
* Migration V10 — additive-only (`suppression_count`, `suppressed_at`
+ partial indices). All v2.0.x DBs upgrade seamlessly on first launch.
* Dashboard: `ForgettingIndicator.svelte` pulses when suppressions
are active. 3D graph nodes dim to 20% opacity when suppressed.
New WebSocket events: `MemorySuppressed`, `MemoryUnsuppressed`,
`Rac1CascadeSwept`. Heartbeat carries `suppressed_count`.
* Search pipeline: SIF penalty inserted into the accessibility stage
so it stacks on top of passive FSRS decay.
* Tool count bumped 23 → 24. Cognitive modules 29 → 30.
Memories persist — they are INHIBITED, not erased. `memory.get(id)`
returns full content through any number of suppressions. The 24h
labile window is a grace period for regret.
Also fixes issue #31 (dashboard graph view buggy) as a companion UI
bug discovered during the v2.0.5 audit cycle:
* Root cause: node glow `SpriteMaterial` had no `map`, so
`THREE.Sprite` rendered as a solid-coloured 1×1 plane. Additive
blending + `UnrealBloomPass(0.8, 0.4, 0.85)` amplified the square
edges into hard-edged glowing cubes.
* Fix: shared 128×128 radial-gradient `CanvasTexture` singleton used
as the sprite map. Retuned bloom to `(0.55, 0.6, 0.2)`. Halved fog
density (0.008 → 0.0035). Edges bumped from dark navy `0x4a4a7a`
to brand violet `0x8b5cf6` with higher opacity. Added explicit
`scene.background` and a 2000-point starfield for depth.
* 21 regression tests added in `ui-fixes.test.ts` locking every
invariant in (shared texture singleton, depthWrite:false, scale
×6, bloom magic numbers via source regex, starfield presence).
Tests: 1,284 Rust (+47) + 171 Vitest (+21) = 1,455 total, 0 failed
Clippy: clean across all targets, zero warnings
Release binary: 22.6MB, `cargo build --release -p vestige-mcp` green
Versions: workspace aligned at 2.0.5 across all 6 crates/packages
Closes #31
2026-04-14 17:30:30 -05:00
let args = serde_json ::json! ( { " id " : " 00000000-0000-0000-0000-000000000000 " } ) ;
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
let result = execute_promote ( & storage , & test_cognitive ( ) , Some ( args ) ) . await ;
assert! ( result . is_err ( ) ) ;
assert! ( result . unwrap_err ( ) . contains ( " Node not found " ) ) ;
}
#[ tokio::test ]
async fn test_promote_missing_id_field_fails ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
let args = serde_json ::json! ( { " reason " : " test " } ) ;
let result = execute_promote ( & storage , & test_cognitive ( ) , Some ( args ) ) . await ;
assert! ( result . is_err ( ) ) ;
assert! ( result . unwrap_err ( ) . contains ( " Invalid arguments " ) ) ;
}
#[ tokio::test ]
async fn test_promote_succeeds ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
let id = ingest_test_memory ( & storage ) . await ;
let args = serde_json ::json! ( { " id " : id , " reason " : " It was helpful " } ) ;
let result = execute_promote ( & storage , & test_cognitive ( ) , Some ( args ) ) . await ;
assert! ( result . is_ok ( ) ) ;
let value = result . unwrap ( ) ;
assert_eq! ( value [ " success " ] , true ) ;
assert_eq! ( value [ " action " ] , " promoted " ) ;
assert_eq! ( value [ " nodeId " ] , id ) ;
assert_eq! ( value [ " reason " ] , " It was helpful " ) ;
assert! ( value [ " changes " ] [ " retrievalStrength " ] . is_object ( ) ) ;
}
#[ tokio::test ]
async fn test_promote_without_reason_succeeds ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
let id = ingest_test_memory ( & storage ) . await ;
let args = serde_json ::json! ( { " id " : id } ) ;
let result = execute_promote ( & storage , & test_cognitive ( ) , Some ( args ) ) . await ;
assert! ( result . is_ok ( ) ) ;
let value = result . unwrap ( ) ;
assert_eq! ( value [ " success " ] , true ) ;
assert! ( value [ " reason " ] . is_null ( ) ) ;
}
#[ tokio::test ]
async fn test_promote_changes_contain_expected_fields ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
let id = ingest_test_memory ( & storage ) . await ;
let args = serde_json ::json! ( { " id " : id } ) ;
let result = execute_promote ( & storage , & test_cognitive ( ) , Some ( args ) ) . await ;
let value = result . unwrap ( ) ;
// Verify response structure includes before/after/delta for all 3 metrics
assert! ( value [ " changes " ] [ " retrievalStrength " ] [ " before " ] . is_number ( ) ) ;
assert! ( value [ " changes " ] [ " retrievalStrength " ] [ " after " ] . is_number ( ) ) ;
assert_eq! ( value [ " changes " ] [ " retrievalStrength " ] [ " delta " ] , " +0.20 " ) ;
assert! ( value [ " changes " ] [ " retentionStrength " ] [ " before " ] . is_number ( ) ) ;
assert! ( value [ " changes " ] [ " retentionStrength " ] [ " after " ] . is_number ( ) ) ;
assert_eq! ( value [ " changes " ] [ " retentionStrength " ] [ " delta " ] , " +0.10 " ) ;
assert! ( value [ " changes " ] [ " stability " ] [ " before " ] . is_number ( ) ) ;
assert! ( value [ " changes " ] [ " stability " ] [ " after " ] . is_number ( ) ) ;
assert_eq! ( value [ " changes " ] [ " stability " ] [ " multiplier " ] , " 1.5x " ) ;
}
// === DEMOTE TESTS ===
#[ tokio::test ]
async fn test_demote_missing_args_fails ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
let result = execute_demote ( & storage , & test_cognitive ( ) , None ) . await ;
assert! ( result . is_err ( ) ) ;
assert! ( result . unwrap_err ( ) . contains ( " Missing arguments " ) ) ;
}
#[ tokio::test ]
async fn test_demote_invalid_uuid_fails ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
let args = serde_json ::json! ( { " id " : " bad-id " } ) ;
let result = execute_demote ( & storage , & test_cognitive ( ) , Some ( args ) ) . await ;
assert! ( result . is_err ( ) ) ;
assert! ( result . unwrap_err ( ) . contains ( " Invalid node ID format " ) ) ;
}
#[ tokio::test ]
async fn test_demote_nonexistent_node_fails ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
feat(v2.0.5): Intentional Amnesia — active forgetting via top-down inhibitory control
First AI memory system to model forgetting as a neuroscience-grounded
PROCESS rather than passive decay. Adds the `suppress` MCP tool (#24),
Rac1 cascade worker, migration V10, and dashboard forgetting indicators.
Based on:
- Anderson, Hanslmayr & Quaegebeur (2025), Nat Rev Neurosci — right
lateral PFC as the domain-general inhibitory controller; SIF
compounds with each stopping attempt.
- Cervantes-Sandoval et al. (2020), Front Cell Neurosci PMC7477079 —
Rac1 GTPase as the active synaptic destabilization mechanism.
What's new:
* `suppress` MCP tool — each call compounds `suppression_count` and
subtracts a `0.15 × count` penalty (saturating at 80%) from
retrieval scores during hybrid search. Distinct from delete
(removes) and demote (one-shot).
* Rac1 cascade worker — background sweep piggybacks the 6h
consolidation loop, walks `memory_connections` edges from
recently-suppressed seeds, applies attenuated FSRS decay to
co-activated neighbors. You don't just forget Jake — you fade
the café, the roommate, the birthday.
* 24h labile window — reversible via `suppress({id, reverse: true})`
within 24 hours. Matches Nader reconsolidation semantics.
* Migration V10 — additive-only (`suppression_count`, `suppressed_at`
+ partial indices). All v2.0.x DBs upgrade seamlessly on first launch.
* Dashboard: `ForgettingIndicator.svelte` pulses when suppressions
are active. 3D graph nodes dim to 20% opacity when suppressed.
New WebSocket events: `MemorySuppressed`, `MemoryUnsuppressed`,
`Rac1CascadeSwept`. Heartbeat carries `suppressed_count`.
* Search pipeline: SIF penalty inserted into the accessibility stage
so it stacks on top of passive FSRS decay.
* Tool count bumped 23 → 24. Cognitive modules 29 → 30.
Memories persist — they are INHIBITED, not erased. `memory.get(id)`
returns full content through any number of suppressions. The 24h
labile window is a grace period for regret.
Also fixes issue #31 (dashboard graph view buggy) as a companion UI
bug discovered during the v2.0.5 audit cycle:
* Root cause: node glow `SpriteMaterial` had no `map`, so
`THREE.Sprite` rendered as a solid-coloured 1×1 plane. Additive
blending + `UnrealBloomPass(0.8, 0.4, 0.85)` amplified the square
edges into hard-edged glowing cubes.
* Fix: shared 128×128 radial-gradient `CanvasTexture` singleton used
as the sprite map. Retuned bloom to `(0.55, 0.6, 0.2)`. Halved fog
density (0.008 → 0.0035). Edges bumped from dark navy `0x4a4a7a`
to brand violet `0x8b5cf6` with higher opacity. Added explicit
`scene.background` and a 2000-point starfield for depth.
* 21 regression tests added in `ui-fixes.test.ts` locking every
invariant in (shared texture singleton, depthWrite:false, scale
×6, bloom magic numbers via source regex, starfield presence).
Tests: 1,284 Rust (+47) + 171 Vitest (+21) = 1,455 total, 0 failed
Clippy: clean across all targets, zero warnings
Release binary: 22.6MB, `cargo build --release -p vestige-mcp` green
Versions: workspace aligned at 2.0.5 across all 6 crates/packages
Closes #31
2026-04-14 17:30:30 -05:00
let args = serde_json ::json! ( { " id " : " 00000000-0000-0000-0000-000000000000 " } ) ;
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
let result = execute_demote ( & storage , & test_cognitive ( ) , Some ( args ) ) . await ;
assert! ( result . is_err ( ) ) ;
assert! ( result . unwrap_err ( ) . contains ( " Node not found " ) ) ;
}
#[ tokio::test ]
async fn test_demote_succeeds ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
let id = ingest_test_memory ( & storage ) . await ;
let args = serde_json ::json! ( { " id " : id , " reason " : " It was wrong " } ) ;
let result = execute_demote ( & storage , & test_cognitive ( ) , Some ( args ) ) . await ;
assert! ( result . is_ok ( ) ) ;
let value = result . unwrap ( ) ;
assert_eq! ( value [ " success " ] , true ) ;
assert_eq! ( value [ " action " ] , " demoted " ) ;
assert_eq! ( value [ " nodeId " ] , id ) ;
assert_eq! ( value [ " reason " ] , " It was wrong " ) ;
assert! ( value [ " note " ] . as_str ( ) . unwrap ( ) . contains ( " NOT deleted " ) ) ;
}
#[ tokio::test ]
async fn test_demote_changes_contain_expected_fields ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
let id = ingest_test_memory ( & storage ) . await ;
let args = serde_json ::json! ( { " id " : id } ) ;
let result = execute_demote ( & storage , & test_cognitive ( ) , Some ( args ) ) . await ;
let value = result . unwrap ( ) ;
assert! ( value [ " changes " ] [ " retrievalStrength " ] [ " before " ] . is_number ( ) ) ;
assert! ( value [ " changes " ] [ " retrievalStrength " ] [ " after " ] . is_number ( ) ) ;
assert_eq! ( value [ " changes " ] [ " retrievalStrength " ] [ " delta " ] , " -0.30 " ) ;
assert_eq! ( value [ " changes " ] [ " retentionStrength " ] [ " delta " ] , " -0.15 " ) ;
assert_eq! ( value [ " changes " ] [ " stability " ] [ " multiplier " ] , " 0.5x " ) ;
}
// === REQUEST FEEDBACK TESTS ===
#[ tokio::test ]
async fn test_request_feedback_missing_args_fails ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
let result = execute_request_feedback ( & storage , None ) . await ;
assert! ( result . is_err ( ) ) ;
}
#[ tokio::test ]
async fn test_request_feedback_invalid_uuid_fails ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
let args = serde_json ::json! ( { " id " : " not-valid " } ) ;
let result = execute_request_feedback ( & storage , Some ( args ) ) . await ;
assert! ( result . is_err ( ) ) ;
}
#[ tokio::test ]
async fn test_request_feedback_nonexistent_node_fails ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
feat(v2.0.5): Intentional Amnesia — active forgetting via top-down inhibitory control
First AI memory system to model forgetting as a neuroscience-grounded
PROCESS rather than passive decay. Adds the `suppress` MCP tool (#24),
Rac1 cascade worker, migration V10, and dashboard forgetting indicators.
Based on:
- Anderson, Hanslmayr & Quaegebeur (2025), Nat Rev Neurosci — right
lateral PFC as the domain-general inhibitory controller; SIF
compounds with each stopping attempt.
- Cervantes-Sandoval et al. (2020), Front Cell Neurosci PMC7477079 —
Rac1 GTPase as the active synaptic destabilization mechanism.
What's new:
* `suppress` MCP tool — each call compounds `suppression_count` and
subtracts a `0.15 × count` penalty (saturating at 80%) from
retrieval scores during hybrid search. Distinct from delete
(removes) and demote (one-shot).
* Rac1 cascade worker — background sweep piggybacks the 6h
consolidation loop, walks `memory_connections` edges from
recently-suppressed seeds, applies attenuated FSRS decay to
co-activated neighbors. You don't just forget Jake — you fade
the café, the roommate, the birthday.
* 24h labile window — reversible via `suppress({id, reverse: true})`
within 24 hours. Matches Nader reconsolidation semantics.
* Migration V10 — additive-only (`suppression_count`, `suppressed_at`
+ partial indices). All v2.0.x DBs upgrade seamlessly on first launch.
* Dashboard: `ForgettingIndicator.svelte` pulses when suppressions
are active. 3D graph nodes dim to 20% opacity when suppressed.
New WebSocket events: `MemorySuppressed`, `MemoryUnsuppressed`,
`Rac1CascadeSwept`. Heartbeat carries `suppressed_count`.
* Search pipeline: SIF penalty inserted into the accessibility stage
so it stacks on top of passive FSRS decay.
* Tool count bumped 23 → 24. Cognitive modules 29 → 30.
Memories persist — they are INHIBITED, not erased. `memory.get(id)`
returns full content through any number of suppressions. The 24h
labile window is a grace period for regret.
Also fixes issue #31 (dashboard graph view buggy) as a companion UI
bug discovered during the v2.0.5 audit cycle:
* Root cause: node glow `SpriteMaterial` had no `map`, so
`THREE.Sprite` rendered as a solid-coloured 1×1 plane. Additive
blending + `UnrealBloomPass(0.8, 0.4, 0.85)` amplified the square
edges into hard-edged glowing cubes.
* Fix: shared 128×128 radial-gradient `CanvasTexture` singleton used
as the sprite map. Retuned bloom to `(0.55, 0.6, 0.2)`. Halved fog
density (0.008 → 0.0035). Edges bumped from dark navy `0x4a4a7a`
to brand violet `0x8b5cf6` with higher opacity. Added explicit
`scene.background` and a 2000-point starfield for depth.
* 21 regression tests added in `ui-fixes.test.ts` locking every
invariant in (shared texture singleton, depthWrite:false, scale
×6, bloom magic numbers via source regex, starfield presence).
Tests: 1,284 Rust (+47) + 171 Vitest (+21) = 1,455 total, 0 failed
Clippy: clean across all targets, zero warnings
Release binary: 22.6MB, `cargo build --release -p vestige-mcp` green
Versions: workspace aligned at 2.0.5 across all 6 crates/packages
Closes #31
2026-04-14 17:30:30 -05:00
let args = serde_json ::json! ( { " id " : " 00000000-0000-0000-0000-000000000000 " } ) ;
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
let result = execute_request_feedback ( & storage , Some ( args ) ) . await ;
assert! ( result . is_err ( ) ) ;
}
#[ tokio::test ]
async fn test_request_feedback_succeeds ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
let id = ingest_test_memory ( & storage ) . await ;
let args = serde_json ::json! ( { " id " : id , " context " : " debugging " } ) ;
let result = execute_request_feedback ( & storage , Some ( args ) ) . await ;
assert! ( result . is_ok ( ) ) ;
let value = result . unwrap ( ) ;
assert_eq! ( value [ " action " ] , " request_feedback " ) ;
assert_eq! ( value [ " nodeId " ] , id ) ;
assert! ( value [ " memoryPreview " ] . is_string ( ) ) ;
assert! ( value [ " options " ] . is_array ( ) ) ;
assert_eq! ( value [ " options " ] . as_array ( ) . unwrap ( ) . len ( ) , 3 ) ;
assert_eq! ( value [ " context " ] , " debugging " ) ;
}
#[ tokio::test ]
async fn test_request_feedback_truncates_long_content ( ) {
let ( storage , _dir ) = test_storage ( ) . await ;
let long_content = " A " . repeat ( 200 ) ;
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
let node = storage
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
. ingest ( vestige_core ::IngestInput {
content : long_content ,
node_type : " fact " . to_string ( ) ,
source : None ,
sentiment_score : 0.0 ,
sentiment_magnitude : 0.0 ,
tags : vec ! [ ] ,
valid_from : None ,
valid_until : None ,
} )
. unwrap ( ) ;
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
let node_id = node . id . clone ( ) ;
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
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
let args = serde_json ::json! ( { " id " : node_id } ) ;
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
let result = execute_request_feedback ( & storage , Some ( args ) ) . await ;
let value = result . unwrap ( ) ;
let preview = value [ " memoryPreview " ] . as_str ( ) . unwrap ( ) ;
assert! ( preview . ends_with ( " ... " ) ) ;
assert! ( preview . len ( ) < = 103 ) ;
}
}