Memory Benchmark — Development skill for Claude Code
Recall-Benchmark des Memory-Index (read-only, ändert nichts).
How to install Memory Benchmark
Installs to ~/.claude/skills/mvcoertel-lang-vault-kit-memory-benchmark/SKILL.md
mkdir -p ~/.claude/skills/mvcoertel-lang-vault-kit-memory-benchmark && curl -fsSL https://raw.githubusercontent.com/mvcoertel-lang/vault-kit/HEAD/commands/memory-benchmark.md -o ~/.claude/skills/mvcoertel-lang-vault-kit-memory-benchmark/SKILL.md Restart Claude Code, or start a new session, for it to be picked up.
What Memory Benchmark does
description: Recall-Benchmark des Memory-Index (read-only, ändert nichts) argument-hint: [anzahl fragen]
Read-only Eval des Memory-Retrievals. **Nichts ändern.**
- Lies
MEMORY.mdund alle Memory-Dateien in__MEMDIR__/. - Formuliere je Memory-Datei eine realistische Frage (wie ein Nutzer sie stellen würde) plus 2–3 Cross-Fragen, die mehrere Dateien berühren oder bewusst mehrdeutig sind.
- Für jede Frage: entscheide nur anhand der
MEMORY.md-Indexzeilen (nicht der Dat
Alternatives in Development
- /diagnose — Vérifie l'état de l'environnement RTK et suggère des corrections 25.5k ★
- Scope Recall Hermes — Local long-term memory for Hermes, Codex, Claude Code, WorkBuddy and dsh agents: one SQLite store, source-back 438 ★
- Engage.Memory — Cross-engagement pattern memory — recall prior techniques, record confirmed findings, housekeeping 348 ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
V Memory Refresh
(Re)index docs/superpowers prose into the local V-memory cache so recall is current. Incremental by file hash;
Deep Recall
Deep semantic search over the ENTIRE vault (2k+ notes incl. session-log, handoffs, transcripts, graph) via the
Uaf Recall
Find where something is implemented using the project index instead of reading files one by one.
Recall Health
Show total-recall index health — last ingest age, hook fire rate, p95 latency, error count, DB size.
Recall Rebuild
Rebuild the total-recall index from scratch (drops then full re-ingest). Destructive.
Innovate
Draft N novel sparse-attention submissions in one shot — algorithm-innovation focus, all must compile, no benc
Related Agents
Lean Reader
Read-only reader, collector, or verifier stage that runs without CLAUDE.md and the memory index (18.8k fewer t
Codebase Memory Auditor
Bounded-scope graph audit with check_index_coverage and source read/grep fallback.
Cognitive Engineer
Agent cognition and memory systems expert. When to use: memory plugin design, decision-making models, agent ps