memory-engineering

{
  "name": "memory-engineering",
  "description": "Engineer an agent's forgetting, not just its remembering. Four deterministic stdlib scripts implement the four lenses of agent memory: a cost profiler that splits construction from query spend and reports cost per correct answer (construction energy exceeds total query energy across 300 queries in the Stanford characterization); an architecture picker that scores the four paradigm families — long-context, flat RAG, structure-augmented RAG, agentic — disqualifies on hard constraints, names the cost the winning choice makes you pay, and refuses to pick when the top two tie; a density auditor that classifies every record in a real memory directory or JSONL export as FACT / SKILL / LOG / PROSE, finds near-duplicates, and flags stale and time-relative wording; and a forgetting-policy linter that fails any design with no forgetting rule or that auto-merges contradictions. Ships cs-memory-engineer agent, /cs:memory-engineering and /cs:forgetting-audit commands, 4 references citing 7 sources each, and a seven-question forcing worksheet.",
  "version": "2.11.2",
  "author": {
    "name": "Alireza Rezvani",
    "url": "https://alirezarezvani.com"
  },
  "homepage": "https://github.com/alirezarezvani/claude-skills/tree/main/engineering/memory-engineering",
  "repository": "https://github.com/alirezarezvani/claude-skills",
  "license": "MIT",
  "skills": [
    "./skills/memory-engineering"
  ],
  "source": {
    "build_pattern": "Four-lens synthesis (Stanford cost / Microsoft what-to-keep / Anthropic control / Nvidia hardware) + 4 deterministic stdlib scripts, with a blocking forgetting gate. Framing synthesized from 'How to be a Memory Engineer' by @N01ennn (x.com/N01ennn/status/2083971749079581120); every quantitative claim is cited to the primary source instead, and two of the article's paraphrases are corrected in the references (the 47x energy figure is the spread across ten systems, not an accuracy-matched pair; the 97% error-reduction figure is one named customer's reported result, not a general property).",
    "distinct_from": "llm-wiki (maintains one specific markdown vault; this audits and prices any memory system); skillopt-sleep (runs a nightly consolidation loop; this decides whether that loop's output is worth keeping); agent-harness (bounds a task loop; this bounds a store); llm-cost-optimizer (prices inference generally, not the memory write path specifically)"
  }
}