Menü
Use when long-running agent loops are hitting context limits, accumulating large tool outputs, or need ReMe message offload and reload patterns to avoid context rot.
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Basierend auf der SOC-Berufsklassifikation
| name | reme-working-memory |
| description | Use when long-running agent loops are hitting context limits, accumulating large tool outputs, or need ReMe message offload and reload patterns to avoid context rot. |
Use this skill for ReMe's working-memory and context-engineering primitives.
Use this skill when you need to:
Do not use this skill for long-term semantic memory policy, personal/task/tool memory classification, or service startup.
This skill centers on two operation families:
Typical reload tools include grep-style search and targeted file reads over offloaded artifacts.
Use when an agent needs a file-first ReMe workspace as the source of durable setup facts, stable preferences, or reusable daily decision records.
Use when an agent needs a file-first ReMe workspace as the source of durable setup facts, stable preferences, or reusable daily decision records.
Use when you need vector-backed personal memory in ReMe, including preference extraction, personal-memory retrieval, or PersonalMemoryService and ReMeApp integration.
Use when integrating ReMeLight into an agent loop and you need file-based context checks, compaction, tool-result offload, semantic search, or session memory.
Use when starting ReMe over HTTP or MCP, or when embedding ReMeApp directly into an application without changing memory-type semantics.
Use when you need ReMe to learn from prior task trajectories, extract reusable success or failure patterns, or retrieve procedural memory for similar tasks.
