ask-mcp

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

DeepWiki Research

DeepWiki 研究

Query repositories via DeepWiki MCP to answer technical questions with precise, source-backed answers.

通过DeepWiki MCP查询代码仓库，获取有精准源码支撑的技术问题答案。

Workflow

工作流程

Clarify intent
- Recommend plan mode, if it's not activated
- Use
```
AskUserQuestion
```
  to resolve blind spots
- All questions must be resolved before execution
Identify references and skills
- Match task to execution steps below
- Locate relevant documentation and examples
Write plan file (YAML task format)
- Use
```
AskUserQuestion
```
  for anything unclear — never guess or assume
- Identify blockers: permissions, dependencies, unknowns
- Plan must be complete before execution begins
Execute
- Use
```
Task
```
  tool with subagents for parallel research
- Subagents load skills via
```
Skill
```
  tool
- Track progress with
```
TodoWrite
```
When stuck: spawn subagent with
```
Read
```
,
```
Glob
```
,
```
Grep
```
, DeepWiki MCP access and load
```
skills/ask-mcp
```

明确意图
- 如果未激活计划模式，推荐启用该模式
- 使用
```
AskUserQuestion
```
  工具解决信息盲区
- 执行前必须解决所有疑问
识别参考资料与技能
- 将任务与下方的执行步骤匹配
- 定位相关文档与示例
编写计划文件（YAML任务格式）
- 任何不明确的内容都使用
```
AskUserQuestion
```
  工具确认——绝不猜测或假设
- 识别阻碍因素：权限、依赖项、未知内容
- 计划必须完整才能开始执行
执行任务
- 使用
```
Task
```
  工具结合子Agent进行并行研究
- 子Agent通过
```
Skill
```
  工具加载技能
- 使用
```
TodoWrite
```
  跟踪进度
遇到阻塞时：创建具备
```
Read
```
、
```
Glob
```
、
```
Grep
```
、DeepWiki MCP访问权限的子Agent，并加载
```
skills/ask-mcp
```
技能

Execution Steps

执行步骤

1. Read Required Context in Local Repo

1. 读取本地仓库中的必要上下文

Use

/init

skill and index current repository, if you have not already

如果尚未初始化，使用

/init

技能并索引当前代码仓库

2. Identify Question Scope

2. 确定问题范围

Determine the domain:

Programs, client SDKs, architecture, implementation details
Specific components (LightAccount, ValidityProof, CPI, etc.)

明确问题所属领域：

程序、客户端SDK、架构、实现细节
特定组件（LightAccount、ValidityProof、CPI等）

2. Fetch Repository Context

2. 获取仓库上下文

Select the appropriate repository based on question scope:

Light Protocol (compressed accounts, state trees, ZK compression, Light SDK)

mcp__deepwiki__read_wiki_structure("Lightprotocol/light-protocol")
mcp__deepwiki__read_wiki_contents("Lightprotocol/light-protocol")
mcp__deepwiki__ask_question("Lightprotocol/light-protocol", "your question")

Solana Development (programs, accounts, general Solana)

mcp__deepwiki__read_wiki_structure("blueshift-gg/blueshift-dashboard")
mcp__deepwiki__read_wiki_contents("blueshift-gg/blueshift-dashboard")
mcp__deepwiki__ask_question("blueshift-gg/blueshift-dashboard", "your question")

Claude Code (CLI features, commands, configuration)

mcp__deepwiki__read_wiki_structure("anthropics/claude-code")
mcp__deepwiki__read_wiki_contents("anthropics/claude-code")
mcp__deepwiki__ask_question("anthropics/claude-code", "your question")

Agent Skills (skills, subagents, patterns)

mcp__deepwiki__read_wiki_structure("anthropics/skills")
mcp__deepwiki__read_wiki_contents("anthropics/skills")
mcp__deepwiki__ask_question("anthropics/skills", "your question")

根据问题范围选择合适的代码仓库：

Light Protocol（压缩账户、状态树、ZK压缩、Light SDK）

mcp__deepwiki__read_wiki_structure("Lightprotocol/light-protocol")
mcp__deepwiki__read_wiki_contents("Lightprotocol/light-protocol")
mcp__deepwiki__ask_question("Lightprotocol/light-protocol", "your question")

Solana开发（程序、账户、通用Solana相关）

mcp__deepwiki__read_wiki_structure("blueshift-gg/blueshift-dashboard")
mcp__deepwiki__read_wiki_contents("blueshift-gg/blueshift-dashboard")
mcp__deepwiki__ask_question("blueshift-gg/blueshift-dashboard", "your question")

Claude Code（CLI功能、命令、配置）

mcp__deepwiki__read_wiki_structure("anthropics/claude-code")
mcp__deepwiki__read_wiki_contents("anthropics/claude-code")
mcp__deepwiki__ask_question("anthropics/claude-code", "your question")

Agent技能（技能、子Agent、模式）

mcp__deepwiki__read_wiki_structure("anthropics/skills")
mcp__deepwiki__read_wiki_contents("anthropics/skills")
mcp__deepwiki__ask_question("anthropics/skills", "your question")

3. Apply Precision Rules

3. 遵循精准性规则

AVOID:

Vague verbs: "handles", "manages", "processes", "enables", "provides"
Abstract nouns: "operations", "management", "coordination"
-ing endings when a precise verb exists
Marketing language: "comprehensive", "rich", "seamless", "enables", "provides capability"
ZK terminology in user-facing docs: "inclusion proof", "non-inclusion proof" → Instead describe what it does: "prove the address doesn't exist", "prove the account hash exists"
Generic descriptions: "account metadata" instead of "CompressedAccountMeta"

USE:

Exact function/method names
Concrete data structures
Specific operations: "sets discriminator", "creates hash", "nullifies account"
File:line references
"reference" when describing u8 indices that reference accounts in the accounts array (standard Solana terminology)
"CPI" as a verb (e.g., "CPIs the Account Compression Program")
Technical method/function names used as verbs when contextually clear

Describe exactly what happens.

需避免：

模糊动词：如‘处理’、‘管理’、‘实现’、‘提供’等
抽象名词：如‘操作’、‘管理’、‘协调’等
已有精准动词时使用-ing形式的表述
营销类话术：如‘全面的’、‘丰富的’、‘无缝的’、‘实现’、‘提供能力’等
用户文档中使用ZK术语：如‘包含证明’、‘不包含证明’ → 改为描述其功能：‘证明地址不存在’、‘证明账户哈希存在’
通用描述：如用‘账户元数据’代替具体的‘CompressedAccountMeta’

需使用：

精确的函数/方法名称
具体的数据结构
特定操作：如‘设置鉴别符’、‘创建哈希’、‘作废账户’等
文件:行号引用
描述账户数组中u8索引时使用‘引用’（符合Solana标准术语）
将"CPI"用作动词（例如‘向Account Compression Program发起CPI调用’）
上下文明确时，将技术方法/函数名称用作动词

准确描述实际发生的操作。

4. Format Response

4. 格式化响应

Structure answers with:

Technical precision
Code examples with inline comments
Source references (file paths, line numbers)

回答需包含以下结构：

技术精准性
带有行内注释的代码示例
来源引用（文件路径、行号）

Example Output

示例输出

Question: What do you include in instruction data for compressed account creation?

Answer:

问题：创建压缩账户时，指令数据中包含哪些内容？

回答：

Compressed Account Creation Instruction Data

压缩账户创建指令数据

Instruction data contains three components for proof verification and tree coordination.

Components:

rust

pub struct InstructionData {
    proof: ValidityProof,
    address_tree_info: PackedAddressTreeInfo,
    output_state_tree_index: u8,
    // + custom account data
}

1. Non-inclusion proof

```
ValidityProof
```
proves the derived address does not exist in the specified address tree
Client generates via
```
getValidityProof()
```
RPC call

2. Tree location parameters

```
PackedAddressTreeInfo
```
: Specifies which address tree registers the address (u8 index, not pubkey)
```
output_state_tree_index
```
: Specifies which state tree stores the compressed account hash

3. Custom account data

Program-specific fields (e.g.,
```
message: String
```
, user data)

Execution flow:

Address tree proves uniqueness, stores address
State tree stores account hash
ValidityProof cryptographically guarantees address is unused

Why separate trees:

Compressed accounts require client-generated cryptographic proof that address doesn't exist (unlike regular Solana where runtime checks PDA existence). Address trees enforce uniqueness; state trees store account hashes.

Packed structs use

u8

indices to reference accounts in

remaining_accounts

, reducing transaction size.

指令数据包含三个用于证明验证和树协调的组件。

组件：

rust

pub struct InstructionData {
    proof: ValidityProof,
    address_tree_info: PackedAddressTreeInfo,
    output_state_tree_index: u8,
    // + 自定义账户数据
}

1. 非存在性证明

```
ValidityProof
```
用于证明派生地址未存在于指定的地址树中
客户端通过
```
getValidityProof()
```
RPC调用生成该证明

2. 树位置参数

```
PackedAddressTreeInfo
```
：指定注册地址的地址树（使用u8索引，而非公钥）
```
output_state_tree_index
```
：指定存储压缩账户哈希的状态树

3. 自定义账户数据

程序特定字段（例如
```
message: String
```
、用户数据）

执行流程：

地址树验证唯一性并存储地址
状态树存储账户哈希
ValidityProof通过密码学方式保证地址未被使用

为何使用分离的树结构：

压缩账户需要客户端生成密码学证明来验证地址不存在（与常规Solana不同，常规Solana由运行时检查PDA是否存在）。地址树确保唯一性；状态树存储账户哈希。

打包结构体使用u8索引引用

remaining_accounts

中的账户，从而减少交易大小。