brain-memory 0.1.0-beta.2 → 0.1.0-beta.21

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (44) hide show
  1. package/README.md +272 -72
  2. package/assets/icon.png +0 -0
  3. package/assets/icon.svg +27 -0
  4. package/bin/brain.js +111 -0
  5. package/bin/cloud-sync.js +149 -0
  6. package/bin/install.js +217 -242
  7. package/bin/memorize.js +436 -0
  8. package/bin/pin.js +44 -0
  9. package/bin/recall.js +237 -0
  10. package/bin/reinforce.js +187 -0
  11. package/bin/session-start.js +292 -0
  12. package/bin/skill.js +67 -0
  13. package/bin/unpin.js +38 -0
  14. package/commands/brain/consolidate.md +7 -2
  15. package/commands/brain/explore.md +7 -2
  16. package/commands/brain/forget.md +7 -2
  17. package/commands/brain/init.md +8 -12
  18. package/commands/brain/memorize.md +76 -144
  19. package/commands/brain/pin.md +37 -0
  20. package/commands/brain/remember.md +48 -105
  21. package/commands/brain/review.md +7 -2
  22. package/commands/brain/skills.md +60 -0
  23. package/commands/brain/sleep.md +68 -27
  24. package/commands/brain/status.md +19 -5
  25. package/commands/brain/sunshine.md +18 -13
  26. package/commands/brain/sync.md +360 -0
  27. package/commands/brain/unpin.md +26 -0
  28. package/package.json +18 -8
  29. package/prompts/claude.md +115 -12
  30. package/prompts/gemini.md +115 -12
  31. package/prompts/openai.md +113 -10
  32. package/prompts/opencode.md +178 -0
  33. package/src/cloud-sync.js +534 -0
  34. package/src/crypto.js +130 -0
  35. package/src/export-import.js +200 -0
  36. package/src/git-sync.js +444 -0
  37. package/src/index-manager.js +205 -18
  38. package/src/installer.js +384 -0
  39. package/src/pinning.js +124 -0
  40. package/src/scorer.js +8 -2
  41. package/src/skills.js +208 -0
  42. package/src/tfidf.js +408 -0
  43. package/hooks/session-end.md +0 -48
  44. package/hooks/session-start.md +0 -48
package/README.md CHANGED
@@ -1,13 +1,21 @@
1
- # Brain Memory
1
+ <p align="center">
2
+ <img src="assets/icon.png" alt="Brain Memory" width="128" height="128">
3
+ </p>
2
4
 
3
- [![CI](https://github.com/onurkarali/brain/actions/workflows/ci.yml/badge.svg)](https://github.com/onurkarali/brain/actions/workflows/ci.yml)
5
+ <h1 align="center">Brain Memory</h1>
6
+
7
+ <p align="center">
8
+ <a href="https://github.com/omelas-tech/brain/actions/workflows/ci.yml"><img src="https://github.com/omelas-tech/brain/actions/workflows/ci.yml/badge.svg" alt="CI"></a>
9
+ <a href="https://www.npmjs.com/package/brain-memory"><img src="https://img.shields.io/npm/v/brain-memory" alt="npm version"></a>
10
+ <a href="https://github.com/omelas-tech/brain/blob/main/LICENSE"><img src="https://img.shields.io/npm/l/brain-memory" alt="license"></a>
11
+ </p>
4
12
 
5
13
  A hierarchical, file-system-based memory plugin for AI coding agents. Inspired by human neuroscience — memories are organized into deep nested life-domain categories, connected via associative networks, strengthened through spaced recall, and naturally decay over time.
6
14
 
7
- Works with **Claude Code**, **Gemini CLI**, and **OpenAI Codex CLI**.
15
+ Works with **Claude Code**, **Gemini CLI**, **OpenAI Codex CLI**, and **OpenCode**.
8
16
 
9
17
  ```
10
- .brain/
18
+ ~/.brain/
11
19
  ├── professional/
12
20
  │ └── companies/
13
21
  │ └── acme-corp/
@@ -36,6 +44,8 @@ Existing AI memory solutions use flat databases with tag-based retrieval. Brain
36
44
  - **Context-dependent recall** — Memories encoded in a similar context to the current session are scored higher
37
45
  - **Spaced reinforcement** — Memories recalled after longer intervals get bigger boosts, cramming produces diminishing returns
38
46
  - **Cognitive types** — Episodic, semantic, and procedural memories each decay differently, just like in the brain
47
+ - **Always-present knowledge** — Pin critical conventions and preferences so they load every session, bypassing recall and never decaying — the facts your agent must never miss
48
+ - **Procedural skills** — Reusable how-to workflows with progressive disclosure, learned automatically from repeated experience and exportable to your agent's native skills
39
49
  - **On-demand depth** — Subcategories are created as needed, not pre-defined
40
50
  - **Consolidation** — Weak related memories merge into stronger combined knowledge
41
51
  - **Zero dependencies** — Pure file I/O, no databases, no servers, no embeddings required
@@ -43,20 +53,46 @@ Existing AI memory solutions use flat databases with tag-based retrieval. Brain
43
53
  ## Install
44
54
 
45
55
  ```bash
46
- npx brain-memory@beta
56
+ npm install -g brain-memory@beta
57
+ brain
47
58
  ```
48
59
 
49
- The interactive installer asks which runtime(s) to configure (Claude Code, Gemini CLI, OpenAI Codex CLI, or all) and whether to install globally or for the current project.
60
+ The first command installs the package globally — this gives you both the interactive setup wizard and the `brain` CLI (`brain recall`, `brain memorize`, `brain reinforce`, …) that agents rely on for deterministic scoring.
61
+
62
+ The second command runs the setup wizard, which asks which runtime(s) to configure (Claude Code, Gemini CLI, OpenAI Codex CLI, or all) and whether to install globally or for the current project.
50
63
 
51
64
  ### Non-interactive
52
65
 
53
66
  ```bash
54
- npx brain-memory --claude --global # Claude Code, global
55
- npx brain-memory --gemini --local # Gemini CLI, local project
56
- npx brain-memory --codex --global # OpenAI Codex CLI, global
57
- npx brain-memory --all --global # All runtimes, global
67
+ brain --claude --global # Claude Code, global
68
+ brain --gemini --local # Gemini CLI, local project
69
+ brain --codex --global # OpenAI Codex CLI, global
70
+ brain --opencode --global # OpenCode, global
71
+ brain --all --global # All runtimes, global
72
+ ```
73
+
74
+ ### Update
75
+
76
+ ```bash
77
+ npm install -g brain-memory@beta
78
+ brain update
79
+ ```
80
+
81
+ The first command updates the package and the `brain` CLI. The second command refreshes the slash command prompts for your installed runtimes. Target specific runtimes with `--claude`, `--gemini`, `--openai`, or `--all`.
82
+
83
+ > **Upgrading from an older version?** The separate binaries (`brain-recall`, `brain-memorize`, `brain-reinforce`, `brain-cloud`, `brain-memory`) were unified into a single `brain` dispatcher — use `brain recall`, `brain memorize`, `brain reinforce`, `brain cloud <…>` instead. Run `brain update` to refresh your runtime prompts to the new command surface.
84
+
85
+ > **Why not `npx`?** `npx` runs the setup wizard in a temporary directory that is discarded after execution. The `brain` CLI (`brain recall`, `brain memorize`, `brain reinforce`) won't be available in your PATH, which means agents will fall back to less reliable manual file operations. Always use `npm install -g` to ensure everything works correctly.
86
+
87
+ ### Uninstall
88
+
89
+ ```bash
90
+ brain uninstall
91
+ npm uninstall -g brain-memory
58
92
  ```
59
93
 
94
+ The first command removes slash commands and prompt sections from your agent configs. The second removes the package and CLI tools. Your `~/.brain/` directory (memories) is preserved by default. Add `--delete-data` to the uninstall command to remove it too. Use `--yes` to skip confirmation prompts.
95
+
60
96
  ### Manual Install
61
97
 
62
98
  Copy the `commands/brain/` directory to your agent's commands folder:
@@ -85,9 +121,11 @@ Then append the contents of the corresponding prompt file to your agent's instru
85
121
 
86
122
  | Command | Description |
87
123
  |---------|-------------|
88
- | `/brain:init` | Initialize `.brain/` directory structure with default categories |
89
- | `/brain:memorize [topic]` | Store memories from current session context |
124
+ | `/brain:init` | Initialize `~/.brain/` directory structure with default categories |
125
+ | `/brain:memorize [topic] [--sync]` | Store memories from current session context (add `--sync` to auto-push) |
90
126
  | `/brain:remember [query]` | Recall relevant memories with spreading activation and context matching |
127
+ | `/brain:pin [id\|query]` | Pin a memory to the always-present tier — loads every session, never decays |
128
+ | `/brain:unpin [id\|query]` | Remove a memory from the always-present tier |
91
129
  | `/brain:review [scope]` | Spaced repetition review session for due memories |
92
130
  | `/brain:explore [category]` | Browse the brain hierarchy with visual tree view |
93
131
  | `/brain:consolidate [scope]` | Merge related weak memories into stronger combined ones |
@@ -95,6 +133,58 @@ Then append the contents of the corresponding prompt file to your agent's instru
95
133
  | `/brain:sunshine [target]` | Deep forensic erasure — trace and remove all references |
96
134
  | `/brain:sleep [scope]` | Full maintenance cycle — 9 neuroscience-inspired phases |
97
135
  | `/brain:status` | Dashboard with brain health metrics and recommendations |
136
+ | `/brain:skills [list\|show\|add\|use\|remove\|export]` | Manage procedural skills — reusable how-to workflows with progressive disclosure |
137
+ | `/brain:sync [subcommand]` | Sync memories via Git remote or export/import for portability |
138
+
139
+ ## Session Lifecycle
140
+
141
+ Brain Memory works automatically in the background — no commands needed for basic context awareness.
142
+
143
+ ### Session Start
144
+
145
+ When a session begins and `~/.brain/` exists, the agent makes a single `brain session-start` call. This aggregator returns one deterministic, **token-budget-bounded** payload so the brain can never bloat the context window (the budget lives in `~/.brain/config.json`):
146
+
147
+ 1. **Pinned memories** — always-present conventions and preferences, injected verbatim regardless of recall score
148
+ 2. **Skills index** — the name + description of each available procedural skill (~100 tokens each; full instructions are loaded only when a task matches)
149
+ 3. **Context recall** — the top memories relevant to the current project
150
+ 4. **Review queue + low-confidence alerts**, then a brief status line:
151
+
152
+ ```
153
+ 🧠 Brain active — 42 memories loaded (8 in current project context)
154
+ 📋 3 memories due for review — run /brain:review
155
+ ```
156
+
157
+ The agent treats pinned facts as active constraints, notes which skills exist, silently internalizes relevant memories, and references them naturally during the session — no information dump.
158
+
159
+ ### Ambient Session Tracking
160
+
161
+ Throughout the session, the agent maintains a running mental log of notable events — decisions made, things learned, insights realized, significant experiences. This is purely internal awareness with no file writes. It ensures that early-session events aren't forgotten by the time the session ends.
162
+
163
+ ### Periodic Memory Checkpoint
164
+
165
+ Every ~10 substantive interactions (file edits, architecture decisions, debugging breakthroughs), the agent evaluates whether memorizable content has accumulated. If so, it appends a brief reminder to its next response:
166
+
167
+ ```
168
+ 🧠 Notable decisions and learnings this session — /brain:memorize when ready
169
+ ```
170
+
171
+ This never interrupts your flow — it's a one-liner appended to an existing response, at most once per ~10 interactions. The counter resets after you run `/brain:memorize`.
172
+
173
+ ### Session End
174
+
175
+ When a session ends, the agent performs two steps in order:
176
+
177
+ 1. **Saves session context** to `~/.brain/contexts.json` — always, even for trivial sessions. This includes a `notable_unsaved` field listing items you didn't memorize, so future sessions can reference what happened.
178
+
179
+ 2. **Suggests memorization** if the session contained meaningful decisions, learnings, or insights:
180
+
181
+ ```
182
+ 💡 This session contained notable decisions and learnings.
183
+ Would you like to store them as brain memories?
184
+ Run /brain:memorize to capture them before this context is lost.
185
+ ```
186
+
187
+ The agent never auto-memorizes without user consent. Context is saved proactively — the agent doesn't wait for an explicit goodbye signal.
98
188
 
99
189
  ## How It Works
100
190
 
@@ -117,7 +207,7 @@ Create → Store → Decay → Recall → Reinforce → Review → Sleep → Arc
117
207
 
118
208
  ### Neuroscience Foundations
119
209
 
120
- Brain Memory v2 is grounded in peer-reviewed neuroscience research. Here's how each mechanism maps to the brain:
210
+ Brain Memory is grounded in peer-reviewed neuroscience research. Here's how each mechanism maps to the brain:
121
211
 
122
212
  | Brain Mechanism | Implementation |
123
213
  |-----------------|---------------|
@@ -158,6 +248,38 @@ Each memory is also classified by how the brain processes it:
158
248
 
159
249
  During sleep, frequently-recalled episodic memories are **crystallized** into semantic memories — the specific event fades but the lesson persists. This mirrors how humans extract general principles from repeated experiences.
160
250
 
251
+ ### Always-Present Memories (Pinning)
252
+
253
+ Recall is probabilistic — a stored preference only applies if scoring happens to surface it. For the facts an agent must **never** miss (coding conventions, standing decisions, hard constraints), that's not good enough. Pinning fixes it:
254
+
255
+ - **`pinned`** — the memory is injected at **every** session start regardless of recall score, and is **decay-exempt** (it never fades). Scope it `global` (loads everywhere) or `project:<name>` (loads only in that project).
256
+ - **`stable`** — an independent flag that exempts a memory from decay and pruning **without** forcing it to always load — for timeless facts you recall on demand but don't want to fade.
257
+
258
+ ```bash
259
+ brain pin <id> --scope global --priority 1 # or: /brain:pin always use tabs
260
+ ```
261
+
262
+ Pinned and stable memories are skipped by sleep-cycle homeostasis and pruning, and `/brain:memorize` proactively proposes pinning durable conventions. The always-present set is budget-capped (`pin_budget_tokens`) so it can't crowd out the context window. This maps cleanly to the **semantic memory** type in the [CoALA](https://arxiv.org/abs/2309.02427) agent-memory model — knowledge that is always in context — decomposed into two orthogonal properties (always-loaded vs. non-decaying).
263
+
264
+ ### Procedural Skills
265
+
266
+ Procedural memory is *how* to do things — reusable, step-by-step workflows stored as `~/.brain/_skills/<name>/SKILL.md`. To keep them from flooding the context window, skills use **three-level progressive disclosure**:
267
+
268
+ | Level | When | What loads |
269
+ |-------|------|------------|
270
+ | **L0** | Every session start | Only each skill's name + description (~100 tokens) |
271
+ | **L1** | A task matches a skill | The full `SKILL.md` step-by-step instructions |
272
+ | **L2** | A step needs them | Referenced `resources/` (templates, scripts) |
273
+
274
+ ```bash
275
+ brain skill list # advertised skills (L0)
276
+ brain skill show structured-code-review # full instructions (L1)
277
+ brain skill use structured-code-review # record outcome (--failed weakens it)
278
+ brain skill export structured-code-review --target claude # → native .claude/skills/
279
+ ```
280
+
281
+ Skills strengthen on successful `use` and weaken on `--failed` — one that fails too often demotes itself out of the advertised L0 index. And they aren't only hand-authored: during sleep, Brain **crystallizes** recurring task-solving patterns from repeated experience into new skills (user-confirmed). Finally, `brain skill export` emits a skill in your agent's **native** format so it becomes directly executable.
282
+
161
283
  ### Memory File Format
162
284
 
163
285
  Each memory is a Markdown file with YAML frontmatter:
@@ -226,8 +348,6 @@ score = 0.38 * relevance
226
348
  - **context_match** (0.14) — How similar the encoding context is to the current session
227
349
  - **salience** (0.08) — Emotional/motivational significance
228
350
 
229
- The formula gracefully degrades for v1 memories missing newer fields — weights are renormalized across available terms.
230
-
231
351
  The agent then decides the response strategy:
232
352
  - **Single strong match** (top score > 0.7) → Return the full memory
233
353
  - **Multiple related** (2-5 candidates > 0.4) → Synthesize a consolidated response
@@ -236,7 +356,7 @@ The agent then decides the response strategy:
236
356
 
237
357
  ### Associative Network
238
358
 
239
- Memories are connected via weighted edges in `.brain/associations.json`:
359
+ Memories are connected via weighted edges in `~/.brain/associations.json`:
240
360
 
241
361
  ```json
242
362
  {
@@ -261,7 +381,7 @@ Memories are connected via weighted edges in `.brain/associations.json`:
261
381
 
262
382
  ### Spaced Reinforcement
263
383
 
264
- Unlike the flat +0.05 boost per recall in v1, spaced reinforcement rewards optimal recall timing:
384
+ Spaced reinforcement rewards optimal recall timing:
265
385
 
266
386
  ```
267
387
  spacingMultiplier = min(3.0, 1.0 + log2(1 + daysSinceLastAccess))
@@ -328,24 +448,52 @@ Each expertise area gets an `_expertise.md` profile documenting what you know we
328
448
 
329
449
  Failed recalls reset the interval to 1 day. Successful recalls extend the interval by the ease factor. This ensures you spend time on memories that need reinforcement, not ones you already know well.
330
450
 
331
- ### Migration from v1
451
+ ### Cross-Agent Memory Sharing
452
+
453
+ `~/.brain/` is a single global directory in the user's home folder. All memories are shared across every project and every supported agent automatically. A decision stored by Claude Code in one project is immediately available to Gemini CLI, OpenAI Codex CLI, or OpenCode in any other project — no configuration, no export, no per-project setup. The format is agent-agnostic: plain Markdown files with YAML frontmatter, readable by any tool.
454
+
455
+ Brain Memory is **model-agnostic** as well as agent-portable: because memory is plain files rather than embeddings welded to a particular model, the LLM underneath your agent can be anything — GPT, Claude, Gemini, or any model routed through a gateway — and your memory is unaffected. **Switch your model or switch your agent, and you keep remembering and pick up exactly where you left off.**
456
+
457
+ To share memories across different machines, use `/brain:sync` (see below).
458
+
459
+ ### Portable Sync
460
+
461
+ `/brain:sync` provides two ways to sync memories across devices — no OAuth apps, no cloud provider setup.
462
+
463
+ **Git remote** — Push/pull `~/.brain/` to any private Git repository. Works with GitHub, GitLab, Codeberg, or any self-hosted Git server. Uses your existing Git/SSH authentication — no additional credentials needed.
332
464
 
333
- Existing v1 brains upgrade automatically. The first `/brain:sleep` after upgrade runs a migration phase that:
334
- - Creates `associations.json` from existing `related` fields
335
- - Backfills `recall_history` from `access_count`
336
- - Sets sensible defaults for new fields (`cognitive_type: semantic`, `salience: 0.5`, `confidence: 0.8`)
337
- - Bumps `index.json` version to 2
465
+ **Export/Import** — Pack the entire `~/.brain/` into a single encrypted file for manual transfer via USB, email, or any file-sharing service.
338
466
 
339
- No data is lost. All new scoring gracefully falls back when newer fields are absent.
467
+ **Key features:**
468
+ - **Manual push/pull** — No background watchers, no auto-sync. You control when data moves.
469
+ - **Optional encryption** — AES-256-GCM encryption with a user-provided passphrase. When enabled, all files are encrypted before committing or exporting.
470
+ - **Merge mode** — Import supports merge mode (only import newer files) or overwrite mode.
471
+ - **Zero dependencies** — Uses Node.js built-in `crypto` and the system `git` binary.
472
+
473
+ **Setup:**
474
+ 1. Create a private Git repo (e.g., `gh repo create brain-data --private`)
475
+ 2. Run `/brain:sync setup git@github.com:you/brain-data.git`
476
+ 3. Use `/brain:sync push` and `/brain:sync pull` to keep memories in sync
477
+
478
+ For one-off transfers, use `/brain:sync export` and `/brain:sync import <path>`.
479
+
480
+ Sync state is stored locally in `~/.brain/.sync/` and is never pushed to the remote.
340
481
 
341
482
  ## File Structure
342
483
 
343
484
  ```
344
- .brain/
485
+ ~/.brain/
345
486
  ├── index.json # Memory inventory — fast lookup for all memories
487
+ ├── config.json # Working-memory token budgets (session-start injection)
346
488
  ├── associations.json # Weighted associative network between memories
347
489
  ├── contexts.json # Session context snapshots for context-dependent recall
348
490
  ├── review-queue.json # Spaced repetition scheduling
491
+ ├── pinned.json # Always-present tier manifest (pinned memory IDs + scope)
492
+ ├── skills-index.json # Advertised procedural skills (L0 — name + description)
493
+ ├── _skills/ # Procedural skills
494
+ │ └── <skill-name>/
495
+ │ ├── SKILL.md # Advertised description + step-by-step instructions
496
+ │ └── resources/ # Optional templates/scripts (loaded only at execution)
349
497
  ├── professional/ # Work, career, technical skills
350
498
  │ ├── _meta.json # Category metadata and stats
351
499
  │ ├── _expertise.md # Generated expertise profile
@@ -368,16 +516,19 @@ No data is lost. All new scoring gracefully falls back when newer fields are abs
368
516
  │ └── events/
369
517
  ├── _consolidated/ # Merged memories from consolidation
370
518
  │ └── _meta.json
371
- └── _archived/ # Decayed memories (recoverable + searchable)
372
- ├── _meta.json
373
- └── index.json # Searchable archive index
519
+ ├── _archived/ # Decayed memories (recoverable + searchable)
520
+ │ ├── _meta.json
521
+ │ └── index.json # Searchable archive index
522
+ └── .sync/ # Sync state (local only, never pushed)
523
+ ├── config.json # Remote URL, encryption flag
524
+ └── repo/ # Hidden git repo for sync
374
525
  ```
375
526
 
376
527
  Subdirectories are created **on demand** — the agent decides placement depth based on how specific the memory is. A generic career thought lands in `professional/`, but a specific deployment incident goes to `professional/companies/acme/projects/alpha/`.
377
528
 
378
529
  ## Configuration
379
530
 
380
- Brain configuration lives in `.brain/index.json` under the `config` key:
531
+ Brain configuration lives in `~/.brain/index.json` under the `config` key:
381
532
 
382
533
  ```json
383
534
  {
@@ -413,53 +564,102 @@ Brain configuration lives in `.brain/index.json` under the `config` key:
413
564
  | `association_config.spreading_activation_depth` | 2 | Maximum hops for spreading activation traversal |
414
565
  | `association_config.spreading_activation_decay` | 0.5 | Decay factor per hop during spreading activation |
415
566
 
416
- ## Project Structure
567
+ ### Working-Memory Budget
568
+
569
+ A separate `~/.brain/config.json` (created lazily with safe defaults) caps how much the brain injects into the context window at session start, so the always-present tier can never crowd out your actual work:
417
570
 
571
+ ```json
572
+ {
573
+ "working_memory_budget_tokens": 3000,
574
+ "pin_budget_tokens": 1500,
575
+ "skills_index_budget_tokens": 800,
576
+ "recall_budget_tokens": 700
577
+ }
418
578
  ```
419
- brain/
420
- ├── bin/
421
- │ └── install.js # Interactive installer (npx brain-memory)
422
- ├── commands/
423
- │ └── brain/ # Slash command definitions
424
- │ ├── init.md
425
- │ ├── memorize.md
426
- │ ├── remember.md
427
- │ ├── review.md
428
- │ ├── explore.md
429
- │ ├── consolidate.md
430
- │ ├── forget.md
431
- │ ├── sunshine.md
432
- │ ├── sleep.md
433
- │ └── status.md
434
- ├── prompts/
435
- │ ├── claude.md # CLAUDE.md content (injected by installer)
436
- │ ├── gemini.md # GEMINI.md content (injected by installer)
437
- │ └── openai.md # AGENTS.md content (injected by installer)
438
- ├── hooks/
439
- │ ├── session-start.md # Ambient memory loading + review notifications
440
- │ └── session-end.md # Auto-memorize suggestion + context capture
441
- ├── templates/
442
- │ └── default-categories.json # Default brain category definitions
443
- ├── src/
444
- │ ├── scorer.js # Decay, spreading activation, context matching, spaced reinforcement
445
- │ └── index-manager.js # Index, associations, contexts, review queue, archive CRUD
446
- ├── CLAUDE.md # Development guide for this repo
447
- ├── package.json
448
- └── README.md
579
+
580
+ | Setting | Default | Description |
581
+ |---------|---------|-------------|
582
+ | `working_memory_budget_tokens` | 3000 | Total token ceiling for the whole `brain session-start` payload |
583
+ | `pin_budget_tokens` | 1500 | Sub-budget for pinned, always-present memories |
584
+ | `skills_index_budget_tokens` | 800 | Sub-budget for the advertised skills index (L0) |
585
+ | `recall_budget_tokens` | 700 | Sub-budget for context-relevant recalled memories |
586
+
587
+ Token counts use a dependency-free heuristic stored on each memory at write time. When pins exceed their budget they're selected by `--priority` then strength, and the overflow is reported rather than silently dropped.
588
+
589
+ ## Benchmark
590
+
591
+ Brain Memory ships with a controlled benchmark suite grounded in the 2025-2026 SOTA in long-term-memory evaluation. The system itself is a direct implementation of the [**CoALA**](https://arxiv.org/abs/2309.02427) agent-memory model (Sumers et al., 2023) — Pinned Tier maps to CoALA's *semantic* memory, procedural skills to *procedural* memory.
592
+
593
+ **Six scenarios**, each describable in one sentence:
594
+
595
+ | Id | Pitch | Tests |
596
+ |---|---|---|
597
+ | **A** *Noisy Project Folder* | "Your brain has 200 memories from 6 projects — does it find the 3 relevant ones?" | Retrieval under distractors (LongMemEval-S analog) |
598
+ | **B** *Three Sessions, One Decision* | "Postgres Monday, gRPC rewrite Wednesday, new resource Friday — still Postgres?" | Multi-session continuity + Pinned Tier ablation |
599
+ | **C** *The Contradiction Test* | "Tabs, then spaces, then tabs again — which version wins?" | Decay-weighted recency + contradiction handling |
600
+ | **D** *Skill Progressive Disclosure* | "Five skills indexed, one needed — does brain load just the one?" | CoALA Phase-2 L0/L1/L2 token efficiency |
601
+ | **E** *Continual Coding* | "Five bugs in order — does bug 5 finish faster than bug 1?" | Forward transfer; agent writes memories between tasks |
602
+ | **F** *Abstention* | "No deployment target in memory — does the agent ask or invent?" | Confabulation resistance |
603
+
604
+ **Methodology highlights** — cross-family LLM judge (Claude judges Gemini and vice-versa) with explicit rubric and position-swap; deterministic 200-memory distractor haystack; real `brain session-start` / `brain recall` integration; tokens-per-successful-task as the headline efficiency metric.
605
+
606
+ > Preliminary smoke results (1 run × Scenario A × Gemini Flash, May 2026): `brain-real` resolves the task at **27.8K tokens-per-success**, vs **50.8K** for naïve context-dump and **86.3K** for brain with pinning disabled — a ~3× token economy improvement attributable directly to the Pinned Tier. Full multi-run results are in progress.
607
+
608
+ Run the benchmarks yourself:
609
+
610
+ ```bash
611
+ cd benchmark
612
+ cp .env.example .env # Add your API keys
613
+ npm test # Unit tests
614
+ node harness/runner.js # Full benchmark (cloud APIs)
449
615
  ```
450
616
 
451
- ## Inspired By
452
-
453
- - [get-shit-done](https://github.com/gsd-build/get-shit-done) — Structured file-based workflow orchestration for AI agents
454
- - [Memory in the Age of AI Agents](https://arxiv.org/abs/2512.13564) — Comprehensive survey on agent memory architectures
455
- - [Mem0](https://arxiv.org/pdf/2504.19413) — Human-like memory reinforcement and decay
456
- - [MemOS](https://arxiv.org/pdf/2507.03724) — Memory lifecycle state management
457
- - [Ebbinghaus forgetting curve](https://en.wikipedia.org/wiki/Forgetting_curve) — Exponential memory decay
458
- - [Spreading activation](https://en.wikipedia.org/wiki/Spreading_activation) — Collins & Loftus network model of semantic memory
459
- - [Hebbian theory](https://en.wikipedia.org/wiki/Hebbian_theory) — "Neurons that fire together wire together"
460
- - [Synaptic homeostasis hypothesis](https://doi.org/10.1016/j.neuron.2013.10.024) — Tononi & Cirelli's theory of sleep function
461
- - [SM-2 algorithm](https://en.wikipedia.org/wiki/SuperMemo#Description_of_SM-2_algorithm) — Spaced repetition scheduling
462
- - [Memory reconsolidation](https://en.wikipedia.org/wiki/Memory_consolidation#Reconsolidation) — How recalling memories makes them temporarily malleable
617
+ Live methodology and per-scenario detail: [brainmemory.ai/docs/benchmarks](https://brainmemory.ai/docs/benchmarks) · Source: [`benchmark/`](benchmark/)
618
+
619
+ ## Contributing
620
+
621
+ Contributions are welcome! See [CONTRIBUTING.md](CONTRIBUTING.md) for development setup, project structure, and how to submit changes.
622
+
623
+ ## References
624
+
625
+ Brain Memory's architecture and benchmark methodology are grounded in the following work.
626
+
627
+ ### Foundations — the agent-memory model Brain implements
628
+
629
+ - [**CoALA — Cognitive Architectures for Language Agents**](https://arxiv.org/abs/2309.02427) (arxiv 2309.02427) — Sumers, Yao, Narasimhan, Griffiths. The agent-memory taxonomy Brain implements directly. Pinned Tier → semantic memory, Skills → procedural memory, session-start aggregator → working memory.
630
+ - [**MemGPT — LLMs as Operating Systems**](https://arxiv.org/abs/2310.08560) (arxiv 2310.08560) — Packer et al. Paging-style memory management.
631
+ - [**Generative Agents — Interactive Simulacra of Human Behavior**](https://arxiv.org/abs/2304.03442) (arxiv 2304.03442) — Park et al. Recency · importance · relevance retrieval blend.
632
+ - [**Memory in the Age of AI Agents**](https://arxiv.org/abs/2512.13564) (arxiv 2512.13564) — Comprehensive survey on agent memory architectures.
633
+ - [**Mem0**](https://arxiv.org/abs/2504.19413) (arxiv 2504.19413) — Human-like memory reinforcement and decay.
634
+ - [**MemOS**](https://arxiv.org/abs/2507.03724) (arxiv 2507.03724) — Memory lifecycle state management.
635
+
636
+ ### Memory benchmarks (the suite this work follows)
637
+
638
+ - [**LongMemEval**](https://arxiv.org/abs/2410.10813) (arxiv 2410.10813) — distractor haystacks (S / M / Oracle), abstention category, GPT-4o judge with 97% human agreement.
639
+ - [**MemoryAgentBench**](https://arxiv.org/abs/2507.05257) (arxiv 2507.05257) — four-competency framework; FactConsolidation inspired Scenario C.
640
+ - [**SWE-Bench-CL**](https://arxiv.org/abs/2507.00014) (arxiv 2507.00014) — repo-scoped chronological evaluation; template for Scenario E.
641
+ - [**LoCoMo**](https://arxiv.org/abs/2402.17753) (arxiv 2402.17753) — long-conversation memory benchmark.
642
+ - [**MIRIX**](https://arxiv.org/abs/2507.07957) (arxiv 2507.07957) — realistic synthetic-but-grounded memory benchmarks.
643
+
644
+ ### Methodology — judging and benchmark hygiene
645
+
646
+ - [**Preference Leakage in LLM-as-judge**](https://arxiv.org/abs/2502.01534) (arxiv 2502.01534) — drives cross-family judging.
647
+ - [**When Judgment Becomes Noise — position bias**](https://arxiv.org/abs/2509.20293) (arxiv 2509.20293) — drives position-swap mitigation.
648
+ - [**Silent Judge — shortcut bias**](https://arxiv.org/abs/2509.26072) (arxiv 2509.26072) — drives rubric-based judging.
649
+ - [**LastingBench**](https://arxiv.org/abs/2506.21614) (arxiv 2506.21614) — benchmark-leakage defense.
650
+
651
+ ### Neuroscience
652
+
653
+ - [**Ebbinghaus forgetting curve**](https://en.wikipedia.org/wiki/Forgetting_curve) — exponential memory decay.
654
+ - [**Spreading activation**](https://en.wikipedia.org/wiki/Spreading_activation) — Collins & Loftus network model of semantic memory.
655
+ - [**Hebbian theory**](https://en.wikipedia.org/wiki/Hebbian_theory) — "neurons that fire together wire together."
656
+ - [**Synaptic homeostasis hypothesis**](https://doi.org/10.1016/j.neuron.2013.10.024) — Tononi & Cirelli's theory of sleep function.
657
+ - [**SM-2 algorithm**](https://en.wikipedia.org/wiki/SuperMemo#Description_of_SM-2_algorithm) — spaced repetition scheduling.
658
+ - [**Memory reconsolidation**](https://en.wikipedia.org/wiki/Memory_consolidation#Reconsolidation) — recalling memories makes them temporarily malleable.
659
+
660
+ ### Related projects
661
+
662
+ - [get-shit-done](https://github.com/gsd-build/get-shit-done) — Structured file-based workflow orchestration for AI agents.
463
663
 
464
664
  ## License
465
665
 
Binary file
@@ -0,0 +1,27 @@
1
+ <svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 1024 1024" width="1024" height="1024">
2
+ <defs>
3
+ <radialGradient id="bm-glow" cx="50%" cy="50%" r="50%">
4
+ <stop offset="0%" stop-color="#FFD27A" stop-opacity="0.9"/>
5
+ <stop offset="100%" stop-color="#E8A33D" stop-opacity="0"/>
6
+ </radialGradient>
7
+ </defs>
8
+ <rect x="0" y="0" width="1024" height="1024" rx="224" ry="224" fill="#07090A"/>
9
+ <circle cx="512" cy="512" r="384" fill="url(#bm-glow)" opacity="0.7"/>
10
+ <g stroke="#E8A33D" stroke-width="38" stroke-linecap="round">
11
+ <line x1="512" y1="512" x2="256" y2="256" stroke-opacity="0.4"/>
12
+ <line x1="512" y1="512" x2="800" y2="256" stroke-opacity="0.82"/>
13
+ <line x1="512" y1="512" x2="864" y2="544" stroke-opacity="0.5"/>
14
+ <line x1="512" y1="512" x2="768" y2="800" stroke-opacity="0.75"/>
15
+ <line x1="512" y1="512" x2="256" y2="736" stroke-opacity="0.4"/>
16
+ <line x1="512" y1="512" x2="192" y2="512" stroke-opacity="0.35"/>
17
+ </g>
18
+ <circle cx="256" cy="256" r="51" fill="#9A6E28"/>
19
+ <circle cx="864" cy="544" r="45" fill="#DCA044"/>
20
+ <circle cx="768" cy="800" r="61" fill="#F0B24A"/>
21
+ <circle cx="256" cy="736" r="51" fill="#A9762E"/>
22
+ <circle cx="192" cy="512" r="42" fill="#8A6022"/>
23
+ <circle cx="800" cy="256" r="83" fill="#FFDD8C"/>
24
+ <circle cx="672" cy="368" r="42" fill="#FFF0CC"/>
25
+ <circle cx="512" cy="512" r="118" fill="#FFF2D2"/>
26
+ <circle cx="512" cy="512" r="61" fill="#FFFFFF"/>
27
+ </svg>
package/bin/brain.js ADDED
@@ -0,0 +1,111 @@
1
+ #!/usr/bin/env node
2
+
3
+ /**
4
+ * brain — unified CLI dispatcher for the Brain Memory system.
5
+ *
6
+ * Routes subcommands to their implementations, replacing the former per-binary
7
+ * surface (brain-recall, brain-reinforce, brain-memorize, brain-cloud). Each
8
+ * delegate module still runs on load and remains invokable as `node bin/<file>`,
9
+ * so direct invocation and the existing test suite are unaffected.
10
+ *
11
+ * Usage:
12
+ * brain Interactive installer (also: brain install)
13
+ * brain install|update|uninstall [flags]
14
+ * brain recall "<query>" [--project P] [--task T] [--top N] [--context] [--reindex]
15
+ * brain memorize [--sync] Store memories from a JSON payload on stdin
16
+ * brain reinforce <id> [<id>...] Spaced reinforcement + Hebbian co-retrieval
17
+ * brain cloud <login|logout|push|pull|status>
18
+ * brain --help | brain --version
19
+ */
20
+
21
+ const path = require('path');
22
+
23
+ const VERSION = require('../package.json').version;
24
+
25
+ // Subcommands whose modules read process.argv.slice(2) directly. We strip the
26
+ // leading subcommand token, then require the module (which executes on load).
27
+ const DELEGATED = {
28
+ recall: 'recall.js',
29
+ reinforce: 'reinforce.js',
30
+ memorize: 'memorize.js',
31
+ cloud: 'cloud-sync.js',
32
+ 'session-start': 'session-start.js',
33
+ pin: 'pin.js',
34
+ unpin: 'unpin.js',
35
+ skill: 'skill.js',
36
+ };
37
+
38
+ // install.js finds its positional subcommand from process.argv itself, so it is
39
+ // required without stripping anything. Bare `brain` also routes here, preserving
40
+ // `npx brain-memory` via npm's single-bin fallback.
41
+ const INSTALLER = new Set(['install', 'update', 'uninstall']);
42
+
43
+ const HELP = `🧠 brain — Brain Memory CLI
44
+
45
+ Usage: brain <command> [options]
46
+
47
+ Memory
48
+ recall "<query>" [--project P] [--task T] [--top N]
49
+ Deterministic recall (TF-IDF + decay + spreading)
50
+ recall --context Session-start recall from project context
51
+ recall --reindex Rebuild the search index
52
+ memorize [--sync] Store memories from a JSON payload on stdin
53
+ reinforce <id> [<id>...] Spaced reinforcement + Hebbian co-retrieval
54
+ pin <id> [--scope global|project:<name>] [--priority N]
55
+ Pin a memory into the always-present tier
56
+ unpin <id> Remove a memory from the always-present tier
57
+ skill <list|show|use|add|remove|export>
58
+ Manage procedural skills (progressive disclosure)
59
+ session-start [--project P] [--task T] [--top N]
60
+ Budget-bounded startup payload (agent-invoked)
61
+
62
+ Sync
63
+ cloud <login|logout|push|pull|status>
64
+
65
+ Setup
66
+ install Interactive installer (default when no command)
67
+ update Update an existing installation
68
+ uninstall [--delete-data] Remove the installation
69
+
70
+ Other
71
+ --help, -h Show this help
72
+ --version, -v Show version
73
+ `;
74
+
75
+ function main() {
76
+ const sub = process.argv[2];
77
+
78
+ if (sub === '--version' || sub === '-v') {
79
+ console.log(VERSION);
80
+ return;
81
+ }
82
+
83
+ if (sub === '--help' || sub === '-h' || sub === 'help') {
84
+ console.log(HELP);
85
+ return;
86
+ }
87
+
88
+ // Bare `brain` or an installer subcommand → run the installer.
89
+ if (sub === undefined || INSTALLER.has(sub)) {
90
+ require(path.join(__dirname, 'install.js'));
91
+ return;
92
+ }
93
+
94
+ const file = DELEGATED[sub];
95
+ if (file) {
96
+ process.argv.splice(2, 1); // drop the subcommand token
97
+ const mod = require(path.join(__dirname, file));
98
+ // Modules guarded with `require.main === module` (e.g. session-start) export
99
+ // a main() that we must invoke; legacy modules already ran on require.
100
+ if (mod && typeof mod.main === 'function') {
101
+ mod.main(process.argv.slice(2));
102
+ }
103
+ return;
104
+ }
105
+
106
+ console.error(`Unknown command: ${sub}\n`);
107
+ console.error(HELP);
108
+ process.exit(1);
109
+ }
110
+
111
+ main();