context-manager

Name: context-manager
Author: jackspace

15stars7forks

How to install context-manager?

jackspace/claudeskillz/context-manager

$npx -y skills add jackspace/ClaudeSkillz --skill context-manager

Installs into the current project.

Summary

Manages permanent memory storage for decisions, blockers, context, preferences, and procedures. Use when user says "remember", "save this decision", "what did we decide", "recall", "search memories", "any blockers", or when making important architectural decisions. Provides SDAM

Files · 1

View on GitHub

SKILL.md

1	# Context Manager
2
3	## Purpose
4
5	Permanent external memory system that compensates for SDAM (no episodic memory). Saves and recalls:
6	- DECISION: Architectural choices, tech stack selections, design decisions
7	- BLOCKER: Active obstacles preventing progress
8	- CONTEXT: Project background, requirements, constraints
9	- PREFERENCE: User preferences, coding style, patterns
10	- PROCEDURE: How-to guides, workflows, processes
11	- NOTE: General information, observations, reminders
12
13	For SDAM users: Complete external memory - never forget decisions or context.
14	For ADHD users: Eliminates decision fatigue - past choices automatically recalled.
15	For dyschronometria: All memories time-anchored with explicit timestamps.
16
17	## Activation Triggers
18
19	- User says: "remember", "save this", "don't forget"
20	- User asks: "what did we decide", "recall", "search for"
21	- User mentions: "decision", "blocker", "preference"
22	- Making important architectural decision (proactive save)
23	- Encountering obstacle (proactive blocker tracking)
24
25	## Core Workflow
26
27	### 1. Save Memory
28
29	When user says "remember [something]":
30
31	Step 1: Classify memory type
32	```
33	DECISION: "remember we're using PostgreSQL"
34	BLOCKER: "remember I can't access the API yet"
35	CONTEXT: "remember this is for BOOSTBOX project"
36	PREFERENCE: "remember I prefer functional components"
37	PROCEDURE: "remember how to deploy: npm run build then rsync"
38	NOTE: "remember to update docs after this feature"
39	```
40
41	Step 2: Extract metadata
42	- Content: The actual memory
43	- Tags: Auto-generate from keywords (e.g., "PostgreSQL" → ["database", "postgresql", "backend"])
44	- Project: Infer from current directory or explicit mention
45	- Timestamp: ISO 8601 format
46
47	Step 3: Read current index
48	```bash
49	# Cross-platform: Use $HOME (Linux/macOS) or %USERPROFILE% (Windows)
50	cat ~/.claude-memories/index.json
51	# Windows PowerShell alternative:
52	# Get-Content "$env:USERPROFILE\.claude-memories\index.json"
53	```
54
55	Step 4: Add to index
56	```json
57	{
58	"version": "1.0.0",
59	"created": "2025-10-17T17:45:00Z",
60	"last_updated": "{current_timestamp}",
61	"total_memories": N + 1,
62	"memories_by_type": {
63	"DECISION": X + 1,
64	...
65	},
66	"memories": [
67	{
68	"id": "{uuid}",
69	"type": "DECISION",
70	"content": "Using PostgreSQL as primary database",
71	"timestamp": "{current_timestamp}",
72	"tags": ["database", "postgresql", "backend"],
73	"project": "boostbox",
74	"context": {
75	"file": "{current_file_if_relevant}",
76	"conversation_id": "{if_available}"
77	}
78	},
79	...existing memories
80	],
81	"tags_index": {
82	"database": ["{uuid1}", "{uuid2}"],
83	"postgresql": ["{uuid}"]
84	},
85	"project_index": {
86	"boostbox": ["{uuid1}", "{uuid2}"],
87	"toolhub": ["{uuid3}"]
88	}
89	}
90	```
91
92	Step 5: Create detailed memory file
93	```bash
94	# Save to category-specific directory
95	# Linux/macOS: ~/.claude-memories/decisions/{uuid}.md
96	# Windows: %USERPROFILE%\.claude-memories\decisions\{uuid}.md
97	~/.claude-memories/decisions/{uuid}.md
98	```
99
100	```markdown
101	# DECISION: Using PostgreSQL
102
103	Date: 2025-10-17T17:45:00Z (2 hours ago)
104	Project: BOOSTBOX
105	Tags: database, postgresql, backend
106
107	## Decision
108
109	Using PostgreSQL as primary database instead of MongoDB.
110
111	## Rationale
112
113	{if provided by user or inferred from conversation}
114
115	## Context
116
117	{surrounding conversation context}
118
119	## Related Memories
120
121	{if any related memories found by tag/project match}
122
123	## Last Updated
124
125	2025-10-17T17:45:00Z
126	```
127
128	Step 6: Confirm to user
129	```
130	✅ Remembered: Using PostgreSQL as primary database
131	📁 Saved to: decisions/{uuid}.md
132	🏷️ Tags: database, postgresql, backend
133	📊 Total memories: {N+1}
134	```
135
136	### 2. Recall Memory
137
138	When user asks "what did we decide about [topic]":
139
140	Step 1: Parse query
141	- Extract keywords: "decide" → search DECISION type
142	- Extract topic: "database" → search tags/content
143
144	Step 2: Search index
145	```javascript
146	// Priority order:
147	1. Exact tag match in requested project
148	2. Exact tag match in any project
149	3. Partial content match in requested project
150	4. Partial content match in any project
151
152	// Sort by:
153	1. Relevance (exact match > partial)
154	2. Recency (newer > older)
155	3. Type priority (BLOCKER > DECISION > others)
156	```
157
158	Step 3: Load detailed memory files
159	```bash
160	# For each matching UUID
161	cat ~/.claude-memories/decisions/{uuid}.md
162	# Windows PowerShell:
163	# Get-Content "$env:USERPROFILE\.claude-memories\decisions\{uuid}.md"
164	```
165
166	Step 4: Present results
167	```
168	🔍 Found 3 memories about "database":
169
170	1. DECISION: Using PostgreSQL (2 days ago)
171	📁 Project: BOOSTBOX
172	💡 Using PostgreSQL as primary database instead of MongoDB
173	🔗 decisions/ab

People who liked this also viewed

6 picks

Type

context-manager

15stars7forks

View on GitHub

How to install context-manager?

jackspace/claudeskillz/context-manager

$npx -y skills add jackspace/ClaudeSkillz --skill context-manager

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Context Manager
2
3	## Purpose
4
5	Permanent external memory system that compensates for SDAM (no episodic memory). Saves and recalls:
6	- DECISION: Architectural choices, tech stack selections, design decisions
7	- BLOCKER: Active obstacles preventing progress
8	- CONTEXT: Project background, requirements, constraints
9	- PREFERENCE: User preferences, coding style, patterns
10	- PROCEDURE: How-to guides, workflows, processes
11	- NOTE: General information, observations, reminders
12
13	For SDAM users: Complete external memory - never forget decisions or context.
14	For ADHD users: Eliminates decision fatigue - past choices automatically recalled.
15	For dyschronometria: All memories time-anchored with explicit timestamps.
16
17	## Activation Triggers
18
19	- User says: "remember", "save this", "don't forget"
20	- User asks: "what did we decide", "recall", "search for"
21	- User mentions: "decision", "blocker", "preference"
22	- Making important architectural decision (proactive save)
23	- Encountering obstacle (proactive blocker tracking)
24
25	## Core Workflow
26
27	### 1. Save Memory
28
29	When user says "remember [something]":
30
31	Step 1: Classify memory type
32	```
33	DECISION: "remember we're using PostgreSQL"
34	BLOCKER: "remember I can't access the API yet"
35	CONTEXT: "remember this is for BOOSTBOX project"
36	PREFERENCE: "remember I prefer functional components"
37	PROCEDURE: "remember how to deploy: npm run build then rsync"
38	NOTE: "remember to update docs after this feature"
39	```
40
41	Step 2: Extract metadata
42	- Content: The actual memory
43	- Tags: Auto-generate from keywords (e.g., "PostgreSQL" → ["database", "postgresql", "backend"])
44	- Project: Infer from current directory or explicit mention
45	- Timestamp: ISO 8601 format
46
47	Step 3: Read current index
48	```bash
49	# Cross-platform: Use $HOME (Linux/macOS) or %USERPROFILE% (Windows)
50	cat ~/.claude-memories/index.json
51	# Windows PowerShell alternative:
52	# Get-Content "$env:USERPROFILE\.claude-memories\index.json"
53	```
54
55	Step 4: Add to index
56	```json
57	{
58	"version": "1.0.0",
59	"created": "2025-10-17T17:45:00Z",
60	"last_updated": "{current_timestamp}",
61	"total_memories": N + 1,
62	"memories_by_type": {
63	"DECISION": X + 1,
64	...
65	},
66	"memories": [
67	{
68	"id": "{uuid}",
69	"type": "DECISION",
70	"content": "Using PostgreSQL as primary database",
71	"timestamp": "{current_timestamp}",
72	"tags": ["database", "postgresql", "backend"],
73	"project": "boostbox",
74	"context": {
75	"file": "{current_file_if_relevant}",
76	"conversation_id": "{if_available}"
77	}
78	},
79	...existing memories
80	],
81	"tags_index": {
82	"database": ["{uuid1}", "{uuid2}"],
83	"postgresql": ["{uuid}"]
84	},
85	"project_index": {
86	"boostbox": ["{uuid1}", "{uuid2}"],
87	"toolhub": ["{uuid3}"]
88	}
89	}
90	```
91
92	Step 5: Create detailed memory file
93	```bash
94	# Save to category-specific directory
95	# Linux/macOS: ~/.claude-memories/decisions/{uuid}.md
96	# Windows: %USERPROFILE%\.claude-memories\decisions\{uuid}.md
97	~/.claude-memories/decisions/{uuid}.md
98	```
99
100	```markdown
101	# DECISION: Using PostgreSQL
102
103	Date: 2025-10-17T17:45:00Z (2 hours ago)
104	Project: BOOSTBOX
105	Tags: database, postgresql, backend
106
107	## Decision
108
109	Using PostgreSQL as primary database instead of MongoDB.
110
111	## Rationale
112
113	{if provided by user or inferred from conversation}
114
115	## Context
116
117	{surrounding conversation context}
118
119	## Related Memories
120
121	{if any related memories found by tag/project match}
122
123	## Last Updated
124
125	2025-10-17T17:45:00Z
126	```
127
128	Step 6: Confirm to user
129	```
130	✅ Remembered: Using PostgreSQL as primary database
131	📁 Saved to: decisions/{uuid}.md
132	🏷️ Tags: database, postgresql, backend
133	📊 Total memories: {N+1}
134	```
135
136	### 2. Recall Memory
137
138	When user asks "what did we decide about [topic]":
139
140	Step 1: Parse query
141	- Extract keywords: "decide" → search DECISION type
142	- Extract topic: "database" → search tags/content
143
144	Step 2: Search index
145	```javascript
146	// Priority order:
147	1. Exact tag match in requested project
148	2. Exact tag match in any project
149	3. Partial content match in requested project
150	4. Partial content match in any project
151
152	// Sort by:
153	1. Relevance (exact match > partial)
154	2. Recency (newer > older)
155	3. Type priority (BLOCKER > DECISION > others)
156	```
157
158	Step 3: Load detailed memory files
159	```bash
160	# For each matching UUID
161	cat ~/.claude-memories/decisions/{uuid}.md
162	# Windows PowerShell:
163	# Get-Content "$env:USERPROFILE\.claude-memories\decisions\{uuid}.md"
164	```
165
166	Step 4: Present results
167	```
168	🔍 Found 3 memories about "database":
169
170	1. DECISION: Using PostgreSQL (2 days ago)
171	📁 Project: BOOSTBOX
172	💡 Using PostgreSQL as primary database instead of MongoDB
173	🔗 decisions/ab

People who liked this also viewed

6 picks

Type