remember by yonatangross/orchestkit
npx skills add https://github.com/yonatangross/orchestkit --skill remember将重要的决策、模式或上下文存储在知识图谱中,供未来会话使用。支持追踪成功/失败结果,以构建最佳实践库。
TEXT = "$ARGUMENTS" # 完整的参数字符串,例如:"We use cursor pagination"
FLAG = "$ARGUMENTS[0]" # 第一个标记 —— 检查是否为 --success、--failed、--category、--agent
# 从 $ARGUMENTS[0]、$ARGUMENTS[1] 等解析标志(CC 2.1.59 索引访问)
# 标志之后的剩余标记 = 需要记住的文本
记住技能使用知识图谱作为存储:
mcp__memory__create_entities 和 mcp__memory__create_relations 进行实体和关系存储 - 免费,零配置,始终可用优势:
自动实体提取:
广告位招租
在这里展示您的产品或服务
触达数万 AI 开发者,精准高效
/ork:remember <文本>
/ork:remember --category <类别> <文本>
/ork:remember --success <文本> # 标记为成功模式
/ork:remember --failed <文本> # 标记为反面模式
/ork:remember --success --category <类别> <文本>
# 代理作用域内存
/ork:remember --agent <代理ID> <文本> # 存储在特定代理作用域
/ork:remember --global <文本> # 存储为跨项目最佳实践
| 标志 | 行为 |
|---|---|
| (默认) | 写入图谱 |
--success | 标记为成功模式 |
--failed | 标记为反面模式 |
--category <类别> | 设置类别 |
--agent <代理ID> | 将内存限定在特定代理作用域 |
--global | 存储为跨项目最佳实践 |
decision - 为何选择 X 而非 Y(默认)architecture - 系统设计和模式pattern - 代码规范和标准blocker - 已知问题和变通方案constraint - 限制和要求preference - 用户/团队偏好pagination - 分页策略database - 数据库模式authentication - 认证方法api - API 设计模式frontend - 前端模式performance - 性能优化--success - 效果良好的模式(积极结果)--failed - 导致问题的模式(反面模式)如果未提供任何标志,内存将作为中性(信息性)内容存储。
检查 --success 标志 → 结果:success
检查 --failed 标志 → 结果:failed
检查 --category <类别> 标志
检查 --agent <代理ID> 标志 → agent_id: "ork:{代理ID}"
检查 --global 标志 → 使用全局 user_id
提取需要记住的文本
如果未指定类别,则根据内容自动检测
| 关键词 | 类别 |
|---|---|
| chose, decided, selected | decision |
| architecture, design, system | architecture |
| pattern, convention, style | pattern |
| blocked, issue, bug, workaround | blocker |
| must, cannot, required, constraint | constraint |
| pagination, cursor, offset, page | pagination |
| database, sql, postgres, query | database |
| auth, jwt, oauth, token, session | authentication |
| api, endpoint, rest, graphql | api |
| react, component, frontend, ui | frontend |
| performance, slow, fast, cache | performance |
如果结果是 "failed",则查找:
从文本中提取实体(Technology、Agent、Pattern、Project、AntiPattern),检测关系模式("X uses Y"、"chose X over Y" 等),然后在知识图谱中创建实体和关系。
加载实体提取规则、类型分配、关系模式和图谱创建示例:Read("${CLAUDE_SKILL_DIR}/references/graph-operations.md")
使用适当的模板(成功、反面模式或中性)显示确认信息,展示创建的实体、关系和图谱统计信息。
加载输出模板和示例:Read("${CLAUDE_SKILL_DIR}/references/confirmation-templates.md")
更新 .claude/memory/MEMORY.md 或项目内存文件时:
## Recent Decisions、## Patterns、## Preferencesmemory 技能的"无需权限的文件操作"部分使用 Read("${CLAUDE_SKILL_DIR}/references/<文件>") 按需加载:
| 文件 | 内容 |
|---|---|
category-detection.md | 用于分类记忆的自动检测规则(优先级顺序) |
graph-operations.md | 实体提取、类型分配、关系模式、图谱创建 |
confirmation-templates.md | 输出模板(成功、反面模式、中性)及使用示例 |
ork:memory - 搜索、加载、同步、可视化(读取端操作)每周安装量
93
仓库
GitHub 星标数
134
首次出现
2026年1月22日
安全审计
安装于
gemini-cli85
opencode84
codex84
github-copilot83
cursor81
amp77
Store important decisions, patterns, or context in the knowledge graph for future sessions. Supports tracking success/failure outcomes for building a Best Practice Library.
TEXT = "$ARGUMENTS" # Full argument string, e.g., "We use cursor pagination"
FLAG = "$ARGUMENTS[0]" # First token — check for --success, --failed, --category, --agent
# Parse flags from $ARGUMENTS[0], $ARGUMENTS[1] etc. (CC 2.1.59 indexed access)
# Remaining tokens after flags = the text to remember
The remember skill uses knowledge graph as storage:
mcp__memory__create_entities and mcp__memory__create_relations - FREE, zero-config, always worksBenefits:
Automatic Entity Extraction:
/ork:remember <text>
/ork:remember --category <category> <text>
/ork:remember --success <text> # Mark as successful pattern
/ork:remember --failed <text> # Mark as anti-pattern
/ork:remember --success --category <category> <text>
# Agent-scoped memory
/ork:remember --agent <agent-id> <text> # Store in agent-specific scope
/ork:remember --global <text> # Store as cross-project best practice
| Flag | Behavior |
|---|---|
| (default) | Write to graph |
--success | Mark as successful pattern |
--failed | Mark as anti-pattern |
--category <cat> | Set category |
--agent <agent-id> | Scope memory to a specific agent |
--global | Store as cross-project best practice |
decision - Why we chose X over Y (default)architecture - System design and patternspattern - Code conventions and standardsblocker - Known issues and workaroundsconstraint - Limitations and requirementspreference - User/team preferencespagination - Pagination strategiesdatabase - Database patternsauthentication - Auth approachesapi - API design patterns--success - Pattern that worked well (positive outcome)--failed - Pattern that caused problems (anti-pattern)If neither flag is provided, the memory is stored as neutral (informational).
Check for --success flag → outcome: success
Check for --failed flag → outcome: failed
Check for --category <category> flag
Check for --agent <agent-id> flag → agent_id: "ork:{agent-id}"
Check for --global flag → use global user_id
Extract the text to remember
If no category specified, auto-detect from content
| Keywords | Category |
|---|---|
| chose, decided, selected | decision |
| architecture, design, system | architecture |
| pattern, convention, style | pattern |
| blocked, issue, bug, workaround | blocker |
| must, cannot, required, constraint | constraint |
| pagination, cursor, offset, page | pagination |
| database, sql, postgres, query | database |
| auth, jwt, oauth, token, session | authentication |
| api, endpoint, rest, graphql | api |
| react, component, frontend, ui | frontend |
| performance, slow, fast, cache | performance |
If outcome is "failed", look for:
Extract entities (Technology, Agent, Pattern, Project, AntiPattern) from the text, detect relationship patterns ("X uses Y", "chose X over Y", etc.), then create entities and relations in the knowledge graph.
Load entity extraction rules, type assignment, relationship patterns, and graph creation examples: Read("${CLAUDE_SKILL_DIR}/references/graph-operations.md")
Display confirmation using the appropriate template (success, anti-pattern, or neutral) showing created entities, relations, and graph stats.
Load output templates and examples: Read("${CLAUDE_SKILL_DIR}/references/confirmation-templates.md")
When updating .claude/memory/MEMORY.md or project memory files:
## Recent Decisions, ## Patterns, ## Preferencesmemory skill's "Permission-Free File Operations" section for the full Edit patternLoad on demand with Read("${CLAUDE_SKILL_DIR}/references/<file>"):
| File | Content |
|---|---|
category-detection.md | Auto-detection rules for categorizing memories (priority order) |
graph-operations.md | Entity extraction, type assignment, relationship patterns, graph creation |
confirmation-templates.md | Output templates (success, anti-pattern, neutral) and usage examples |
ork:memory - Search, load, sync, visualize (read-side operations)Weekly Installs
93
Repository
GitHub Stars
134
First Seen
Jan 22, 2026
Security Audits
Gen Agent Trust HubPassSocketPassSnykPass
Installed on
gemini-cli85
opencode84
codex84
github-copilot83
cursor81
amp77
React 组合模式指南:Vercel 组件架构最佳实践,提升代码可维护性
120,000 周安装
frontend - Frontend patternsperformance - Performance optimizations