@rembr/vscode 1.0.0

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.
@@ -0,0 +1,141 @@
1
+ # REMBR - Recursive Language Model with Semantic Memory
2
+
3
+ You have access to REMBR, a semantic memory system via MCP tools. Use it to implement the Recursive Language Model (RLM) pattern for complex tasks.
4
+
5
+ ## Core Principles
6
+
7
+ 1. **Never work with more context than necessary** - Query REMBR for only relevant prior knowledge
8
+ 2. **Break down complex tasks** - Spawn focused subtasks with targeted context
9
+ 3. **Store findings persistently** - Save discoveries to REMBR for future sessions
10
+ 4. **Coordinate through structured returns** - Each subtask returns a standard result format
11
+
12
+ ## Available REMBR Tools (via MCP)
13
+
14
+ ### Memory Management
15
+ - `store_memory` - Store insights, facts, preferences, project info
16
+ - `search_memory` - Hybrid semantic + text search across memories
17
+ - `list_memories` - Browse recent memories by category
18
+ - `get_memory` - Retrieve specific memory by ID
19
+ - `delete_memory` - Remove outdated memories
20
+
21
+ ### Context Workspaces
22
+ - `create_context` - Create workspace for related memories
23
+ - `add_memory_to_context` - Link memories to contexts
24
+ - `search_context` - Scoped search within a workspace
25
+ - `list_contexts` - View available workspaces
26
+
27
+ ### Snapshots (for subtask handoffs)
28
+ - `create_snapshot` - Immutable memory bundle with TTL
29
+ - `get_snapshot` - Retrieve snapshot by ID
30
+ - `list_snapshots` - View available snapshots
31
+
32
+ ### Insights
33
+ - `get_stats` - Usage and limits
34
+ - `get_context_insights` - Category distribution, patterns
35
+ - `get_memory_graph` - Memory relationships
36
+ - `detect_contradictions` - Find conflicting information
37
+
38
+ ## Memory Categories
39
+
40
+ Organize memories semantically:
41
+ - **facts** - Concrete information and data
42
+ - **preferences** - User settings and choices
43
+ - **conversations** - Conversation context
44
+ - **projects** - Project-specific information
45
+ - **learning** - Knowledge and insights
46
+ - **goals** - Objectives and targets
47
+ - **context** - Situational context
48
+ - **reminders** - Future actions
49
+
50
+ ## RLM Pattern: Parent-Subtask Protocol
51
+
52
+ ### Before Decomposing
53
+
54
+ 1. Generate unique `taskId` (e.g., `implement-auth-20240106`)
55
+ 2. Query REMBR for relevant prior context:
56
+ ```
57
+ search_memory({ query: "authentication patterns OAuth JWT", limit: 10 })
58
+ ```
59
+ 3. Identify subtasks and what context each needs
60
+
61
+ ### When Creating Flows/Cascades
62
+
63
+ Provide each flow with:
64
+ ```
65
+ ## Objective
66
+ [Specific focused objective]
67
+
68
+ ## Context from REMBR
69
+ [Relevant memories retrieved for this flow]
70
+
71
+ ## Storage Instructions
72
+ Store findings with:
73
+ - Category: "facts"
74
+ - Metadata: { "taskId": "implement-auth-20240106", "area": "oauth-flow" }
75
+
76
+ ## Return Format
77
+ - Summary: [What you found/did]
78
+ - Findings stored: [Search query to retrieve details]
79
+ - Key points: [Most important items]
80
+ - Status: [complete/partial/blocked]
81
+ ```
82
+
83
+ ### After Flows Complete
84
+
85
+ 1. Read each flow's summary and key points
86
+ 2. If full details needed, query REMBR:
87
+ ```
88
+ search_memory({ query: "implement-auth-20240106 oauth-flow", category: "facts" })
89
+ ```
90
+ 3. Synthesize findings and store to REMBR for future sessions
91
+
92
+ ## Example Usage
93
+
94
+ ### Storing a Discovery
95
+ ```
96
+ store_memory({
97
+ category: "facts",
98
+ content: "API uses rate limiting: 100 req/min per user via express-rate-limit + Redis",
99
+ metadata: {
100
+ taskId: "rate-limit-analysis",
101
+ area: "api-middleware",
102
+ file: "src/middleware/rateLimit.ts"
103
+ }
104
+ })
105
+ ```
106
+
107
+ ### Retrieving Context
108
+ ```
109
+ search_memory({
110
+ query: "rate limiting middleware patterns",
111
+ category: "facts",
112
+ limit: 5
113
+ })
114
+ ```
115
+
116
+ ### Creating a Workspace
117
+ ```
118
+ create_context({
119
+ name: "payment-service-refactor",
120
+ description: "Refactoring payment service to microservices",
121
+ category: "projects"
122
+ })
123
+ ```
124
+
125
+ ## When to Use REMBR
126
+
127
+ ✅ **Do use** for:
128
+ - Storing architectural decisions
129
+ - Remembering project patterns and conventions
130
+ - Building knowledge across sessions
131
+ - Coordinating complex multi-step tasks (Cascade flows)
132
+ - Sharing context between flows
133
+
134
+ ❌ **Don't use** for:
135
+ - Temporary variables or transient data
136
+ - Information already in the codebase
137
+ - Simple single-step tasks
138
+
139
+ ## API Key Setup
140
+
141
+ Ensure `REMBR_API_KEY` is set in your Windsurf environment settings.