Extract session insights into structured Markdown memory files.
npx skills add https://github.com/erclx/aitk --skill claude-memory-capture
Please help me install this Agent Skill. Skill: claude-memory-capture Source: https://github.com/erclx/aitk/tree/main/claude/skills/claude-memory-capture Command: npx skills add https://github.com/erclx/aitk --skill claude-memory-capture
This skill solves the problem of context drift and lost knowledge by automatically extracting valuable patterns, user preferences, and project decisions from active AI sessions into persistent documentation.
Ask the AI to capture the current session memory to update your project documentation.
High-intent search queries and answers about installing and using this skill.
To save Claude Code session context to durable memory, prompt the AI to capture the current session. It scans context for patterns and decisions, then writes them as structured Markdown files into the project's .claude/memory/ directory.
Preventing context drift requires extracting session insights like user preferences and project decisions into persistent Markdown documentation. This ensures consistent AI behavior across future interactions by maintaining a durable knowledge base.
Yes, Claude Code memory capture requires access to the .claude/memory/ directory within the project root. It writes structured Markdown files there, adhering to project-specific prose standards for documentation.
Yes, you can automatically extract user preferences and project decisions by asking the AI to capture session memory. It identifies feedback, decisions, and external references, saving them as rules for future AI interactions.
The process relies on scanning active session context for feedback and decisions, requiring adherence to project-specific prose standards. It operates strictly within the project root directory to maintain structured Markdown documentation.
You need to capture session memory at the end of a complex development session. This ensures specific user preferences or project-wide decisions are saved as durable Markdown rules to guide future AI interactions.
salemazielDefine project directory structures, file naming conventions, and working standards.
jackspaceStore and retrieve persistent memories across sessions with full-text search.
arbgjrParse Claude Code session logs to extract decisions, blockers, and learnings.
haryelramalhoCapture durable project knowledge as categorized Markdown notes with index maintenance.
