Detect contradictions and stale data in stored memories.
npx skills add https://github.com/pokibao/claude-skills-ai-quality --skill memory-audit-pokibao
Please help me install this Agent Skill. Skill: memory-audit Source: https://github.com/pokibao/claude-skills-ai-quality/tree/main/memory-audit Command: npx skills add https://github.com/pokibao/claude-skills-ai-quality --skill memory-audit-pokibao
This Skill detects contradictions and outdated information in stored memories, enabling continuous knowledge refinement to maintain accuracy.
Use the memory-audit skill to scan stored memories for contradictions and outdated entries.
High-intent search queries and answers about installing and using this skill.
Detect contradictions in a knowledge base by applying pattern search and conflict arbitration to stored memories. This process identifies conflicting information and outdated entries to maintain data consistency and integrity for AI context management.
Automate memory analysis by running scheduled audits that scan stored memories for contradictions and outdated entries. Automatic pruning then removes stale or invalid data based on recency and confidence criteria to ensure information accuracy.
Knowledge refinement for AI context management works by extracting high-level rules from scattered memories. This improves core assumptions and continuously refines data quality during idle system time.
Automated memory audits are highly effective for ongoing projects. The system performs memory consistency checks during idle time to prevent conflicting information from accumulating and disrupting project continuity.
AI assistants provide conflicting information when stored memories contain stale data and unresolved contradictions. Applying conflict arbitration and automatic pruning resolves these inconsistencies and restores information integrity.
No external dependencies are required to run memory consistency checks. The skill operates independently using its internal scripts to apply pattern search and conflict arbitration directly to your stored data.
zjykzkIngest raw text data into an Obsidian wiki as searchable knowledge.
Clay-HHKFine-tune large language models with optimized LoRA and QLoRA workflows.
ArchitectVS7Optimize AgentDB vector databases with quantization, HNSW indexing, caching, and batch operations.
superfhpCreate and manage interlinked markdown knowledge bases for research materials.
