Back to Details

braindump

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

braindump - Agent Memory

braindump - Agent 内存

Local, searchable notes that persist across conversations. Silently capture contextual information as longer-term memory that survives beyond the current session.
Manual triggers:
  • "use your brain" or "remember anything about..." — search braindump for relevant context before performing the task
  • "braindump this" or "save this to memory" — store the current context/information for later retrieval
本地可搜索的跨会话留存笔记。将上下文信息静默捕获为可跨当前会话留存的长期内存。
手动触发方式：
  • "use your brain" 或 "remember anything about..." — 在执行任务前搜索braindump中的相关上下文
  • "braindump this" 或 "save this to memory" — 将当前上下文/信息存储以便后续检索

Setup

安装配置

Before first use, verify the CLI is installed. If 
braindump
is not found on PATH, install it:
bash
curl -fsSL https://raw.githubusercontent.com/MohGanji/braindump/main/install.sh | bash
Or run the bundled installer:
bash
bash "SKILL_DIR/scripts/install.sh"
After installation, verify with 
braindump --version
.
首次使用前，请确认CLI已安装。若PATH中未找到
braindump
，请执行以下命令安装：
bash
curl -fsSL https://raw.githubusercontent.com/MohGanji/braindump/main/install.sh | bash
或运行捆绑的安装脚本：
bash
bash "SKILL_DIR/scripts/install.sh"
安装完成后，可通过
braindump --version
验证安装。

Commands

命令

bash
undefined
bash
undefined

Save

保存

braindump add <category> --title "..." --content "..." --tags "tag1,tag2"
braindump add <category> --title "..." --content "..." --tags "tag1,tag2"

Retrieve

检索

braindump search "query" braindump list [category] braindump get <category> "pattern"
braindump search "query" braindump list [category] braindump get <category> "pattern"

Manage

管理

braindump update <id> --content "..." braindump append <id> --content "additional info" braindump delete <id> braindump categories braindump tags

Add `--format json` for programmatic use.
braindump update <id> --content "..." braindump append <id> --content "additional info" braindump delete <id> braindump categories braindump tags

添加`--format json`参数可用于程序化调用。

Autonomous Behavior

自主行为

When to Store (Proactive)

何时存储（主动触发）

Capture information that would be lost when the conversation ends:
  • Business requirements, use cases, user stories
  • Useful URLs and variables (API docs, endpoints, environment configs)
  • API specifications, field mappings, data transformations
  • System constraints, assumptions, exclusions
  • Integration-specific behavior, quirks, gotchas
  • Domain terminology, aliases, abbreviations
  • Technical decisions with rationale
  • Known issues, limitations, workarounds
  • Configuration requirements, thresholds, defaults
  • User preferences and project conventions
  • External context shared by the user (specs, documentation)
  • Resolved bugs and their solutions
捕获会话结束后会丢失的信息：
  • 业务需求、用例、用户故事
  • 有用的URL和变量（API文档、端点、环境配置）
  • API规范、字段映射、数据转换规则
  • 系统约束、假设、排除项
  • 集成特定行为、特性、注意事项
  • 领域术语、别名、缩写
  • 带有决策依据的技术方案
  • 已知问题、限制、解决方法
  • 配置要求、阈值、默认值
  • 用户偏好和项目约定
  • 用户共享的外部上下文（规范、文档）
  • 已修复的Bug及其解决方案

When to Retrieve (Proactive)

何时检索（主动触发）

Search braindump at the start of tasks that might benefit from prior context:
  • Before implementing a feature in a domain previously discussed
  • When the user references something from a past conversation
  • When working with an API, integration, or system previously documented
  • Before making architectural decisions that may have prior rationale recorded
在任务开始时搜索braindump，若过往上下文对当前任务有帮助：
  • 在之前讨论过的领域中实现功能前
  • 当用户引用过往会话中的内容时
  • 处理之前已记录过的API、集成或系统时
  • 做出可能已有过往决策依据的架构决策前

Workflow

工作流程

  1. Before storing: search existing content first — update or append if found, add new note if not
  2. Before working: search for relevant context that may inform the current task
  3. Merge related information under existing categories/titles when possible
  4. Preserve existing content unless contradicted by new information
  5. Focus on evergreen knowledge, not conversation artifacts
  1. 存储前：先搜索现有内容——若找到则更新或追加，未找到则添加新笔记
  2. 工作前：搜索可能为当前任务提供参考的相关上下文
  3. 合并相关信息：尽可能将相关信息归入现有分类/标题下
  4. 保留现有内容：除非新信息与之矛盾，否则不修改现有内容
  5. 聚焦长效知识：不存储会话中的临时产物

Categories and Structure

分类与结构

Categories represent cohesive domain areas: an integration, a system capability, a distinct module, or a logical boundary. Choose categories intuitively based on context — use existing categories when appropriate, create new ones when needed.
Titles should be searchable keywords that narrow context effectively.
Content should be concise, fact-dense paragraphs. Use bullet points for lists. Include code examples only when they clarify behavior.
分类代表连贯的领域范围：一个集成、一项系统能力、一个独立模块或一个逻辑边界。根据上下文直观选择分类——合适时使用现有分类，必要时创建新分类。
标题应采用便于搜索的关键词，有效缩小上下文范围。
内容需简洁、信息密度高的段落。列表使用项目符号。仅当代码示例能阐明行为时才包含。

What NOT to Capture

无需捕获的内容

  • Temporary debugging sessions or transient state
  • File paths or code snippets without context
  • General programming knowledge available in docs
  • Meta-commentary about the conversation itself
  • Information that changes frequently without lasting value
  • 临时调试会话或瞬态状态
  • 无上下文的文件路径或代码片段
  • 文档中可查的通用编程知识
  • 关于会话本身的元评论
  • 频繁变化且无长期价值的信息

Storage

存储位置

~/.braindump/
— Plain text Markdown files with YAML frontmatter. Each category is a directory. An FTS5 SQLite index powers fast search.
~/.braindump/
— 带有YAML前置元数据的纯文本Markdown文件。每个分类对应一个目录。FTS5 SQLite索引支持快速搜索。