Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when working with persistent memory across sessions, storing/retrieving knowledge, managing TODOs/issues, or when context from previous sessions would be helpful.
.claude/skills/agentic-box-memora/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | 23% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 87% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -12% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -42% | 0% |
Memora is the persistent memory system for this environment. Use memora MCP tools to store, search, and organize knowledge across sessions.
memory_create to save architectural decisions, patterns, preferencesmemory_hybrid_search to find relevant past workmemory_create_todo / memory_create_issue for task trackingmemory_hierarchy to browse organized memoriesmemory_create - Store a new memory (auto-deduplicates, suggests hierarchy)memory_create_todo - Create a TODO with priority (high/medium/low)memory_create_issue - Create an issue with severity (critical/major/minor)memory_create_section - Create organizational headersmemory_create_batch - Bulk create multiple memoriesmemory_hybrid_search - Best search: combines keyword + semantic (use this by default)memory_semantic_search - Pure vector similarity searchmemory_list - List/filter by tags, dates, metadatamemory_list_compact - Lightweight listing (id, preview, tags only)memory_hierarchy - View memories in section/subsection treememory_tags - List allowed tagsmemory_tag_hierarchy - View tag namespace treememory_link - Create typed relationships between memoriesmemory_clusters - Detect related memory clustersmemory_find_duplicates - Find and review potential duplicates (LLM-powered)memory_merge - Merge two memories togethermemory_insights - Get activity summary, stale items, patternsmemory_stats - Database statisticsmemory_boost - Increase a memory's importance rankingmemory_export_graph - Export as interactive HTML fileUse hierarchical tags with / separators:
memora/knowledge - General knowledgememora/todos - Task itemsmemora/issues - Bug/issue trackingmemora/auto-capture - Auto-captured contentmemora/sections - Organizational headersproject-name/topic - Project-specific tagssection, subsection, project)When MEMORA_AUTO_CAPTURE=true is set, the PostToolUse hook automatically captures:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-03 | fail→fail | 5,786 | 3,391 | -41% | 1 | 1 | 0% | 839 | 980 | +17% | 0 | 0 | — |
case-04 | pass→pass | 7,289 | 2,002 | -73% | 1 | 1 | 0% | 1,033 | 1,015 | -2% | 0 | 0 | — |
case-01 | fail→fail | 9,388 | 5,896 | -37% | 1 | 1 | 0% | 1,355 | 1,105 | -18% | 0 | 0 | — |
case-02 | fail→fail | 8,102 | 3,717 | -54% | 1 | 1 | 0% | 1,138 | 898 | -21% | 0 | 0 | — |
case-05 | fail→pass | 5,403 | 2,199 | -59% | 1 | 1 | 0% | 887 | 1,088 | +23% | 0 | 0 | — |
case-06 | fail→pass | 5,308 | 4,519 | -15% | 1 | 1 | 0% | 739 | 1,379 | +87% | 0 | 0 | — |
case-07 | fail→pass | 8,789 | 3,790 | -57% | 1 | 1 | 0% | 1,526 | 1,327 | -13% | 0 | 0 | — |
case-08 | fail→pass | 7,257 | 2,297 | -68% | 1 | 1 | 0% | 1,230 | 1,077 | -12% | 0 | 0 | — |
case-09 | fail→fail | 8,947 | 1,958 | -78% | 1 | 1 | 0% | 1,290 | 973 | -25% | 0 | 0 | — |
case-10 | fail→pass | 12,538 | 2,506 | -80% | 1 | 1 | 0% | 1,915 | 1,107 | -42% | 0 | 0 | — |
case-11 | fail→pass | 4,527 | 2,084 | -54% | 1 | 1 | 0% | 753 | 1,025 | +36% | 0 | 0 | — |
case-12 | fail→pass | 11,762 | 2,203 | -81% | 1 | 1 | 0% | 1,812 | 1,011 | -44% | 0 | 0 | — |
case-13 | fail→pass | 6,281 | 1,695 | -73% | 1 | 1 | 0% | 980 | 991 | +1% | 0 | 0 | — |
case-14 | fail→pass | 12,929 | 2,798 | -78% | 1 | 1 | 0% | 2,046 | 1,176 | -43% | 0 | 0 | — |
case-15 | fail→pass | 10,435 | 1,621 | -84% | 1 | 1 | 0% | 1,573 | 935 | -41% | 0 | 0 | — |
case-16 | fail→pass | 14,812 | 1,511 | -90% | 1 | 1 | 0% | 2,229 | 941 | -58% | 0 | 0 | — |
case-17 | fail→pass | 8,251 | 2,673 | -68% | 1 | 1 | 0% | 1,372 | 1,212 | -12% | 0 | 0 | — |
case-18 | fail→pass | 9,572 | 1,585 | -83% | 1 | 1 | 0% | 1,503 | 946 | -37% | 0 | 0 | — |
case-19 | fail→pass | 8,229 | 1,559 | -81% | 1 | 1 | 0% | 1,276 | 923 | -28% | 0 | 0 | — |
case-20 | pass→pass | 3,544 | 2,458 | -31% | 1 | 1 | 0% | 420 | 1,086 | +159% | 0 | 0 | — |
case-21 | pass→pass | 7,857 | 7,690 | -2% | 1 | 1 | 0% | 1,426 | 2,126 | +49% | 0 | 0 | — |
case-22 | pass→pass | 7,711 | 4,921 | -36% | 1 | 1 | 0% | 1,303 | 1,449 | +11% | 0 | 0 | — |
case-23 | pass→pass | 6,031 | 7,851 | +30% | 1 | 1 | 0% | 1,219 | 2,332 | +91% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted, and 20 counted toward the lift figure. The other 3 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +61 percentage points is the difference between those two pass rates over the 20 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.