Agent skill

Common Retrieval

by lihaozheCharlie in lihaozheCharlie/the-way-here

知识库共享检索:绑定资料范围,关键词召回、来源和时间辨析、渐进阅读、双链追溯与证据核验。由查询、预测、了解度和构建 Skill 按需调用;不负责回答格式、未来推演或写入。

MITAuto-check passed

Install Common Retrieval

skills CLI
$ npx skills add lihaozheCharlie/the-way-here --skill common-retrieval -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install lihaozheCharlie/the-way-here common-retrieval --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/lihaozheCharlie/the-way-here.git skills-src && mkdir -p .claude/skills && cp -r skills-src/knowledge-engine/skills/common/retrieval .claude/skills/common-retrieval && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
common-retrieval
GitHub stars
145
Token cost
~399 tokens
SKILL.md length
48 words
Files
5 (incl. scripts, references)
Skills in repo
8
Repo updated
First seen
Licence
MIT

At a glance

知识库共享检索:绑定资料范围,关键词召回、来源和时间辨析、渐进阅读、双链追溯与证据核验。由查询、预测、了解度和构建 Skill 按需调用;不负责回答格式、未来推演或写入。

  • Works in 5 steps: 先找最具体的页面或来源。综合问题通常从相关 Wiki… → 用具体对象、标题、别名、关键词及近义表达召回,命中片段只用于选择阅读位置。标题/别… → 读相关段落及上下文;重要页面读全文。区分已发现的文件和实际读过的内容,截断片段不能… → …
  • SKILL.md covers 绑定资料, 按问题逐步读取, 来源、时间和去重 and 交给调用者
  • Runs Python and TypeScript scripts from its folder

What it does

Common Retrieval is an agent skill from lihaozheCharlie/the-way-here. 知识库共享检索:绑定资料范围,关键词召回、来源和时间辨析、渐进阅读、双链追溯与证据核验。由查询、预测、了解度和构建 Skill 按需调用;不负责回答格式、未来推演或写入。

Its SKILL.md is about 400 tokens, which your agent loads only when the skill is triggered. The skill folder holds 6 other files, including scripts and reference files (for example `references/source-policy.json`, `references/tools.md` and `scripts/evidence_reader.py`).

The repository describes itself as: the-way-here:想成为一个会记得你的来时路、越聊越懂你的朋友。 The licence is MIT.

Example prompts

  • “/common-retrieval”

Requirements

  • Python 3

Workflow steps

5 steps, taken from the first numbered list in SKILL.md.

  1. 先找最具体的页面或来源。综合问题通常从相关 Wiki 开始,原句核验从来源开始,构建从新增来源及对应目标页开始。索引用于定位,不必先读完整总入口。
  2. 用具体对象、标题、别名、关键词及近义表达召回,命中片段只用于选择阅读位置。标题/别名命中也可能相关;不要仅靠文件日期或词频决定重要性。
  3. 读相关段落及上下文;重要页面读全文。区分已发现的文件和实际读过的内容,截断片段不能冒充全文核验。
  4. 围绕缺口沿出链、反链追溯直接来源、同一事件、后续变化、修正和反例。每次扩展有一个具体目的,通常先一跳;泛主题链接和高连接度不自动表示重要。歧义/未解析链接只作为待核查线索。
  5. 对解释、趋势或决策所依赖的关键判断查反例和后续撤回;简单事实定位无需强制多轮。答案已获支持时停止,连续两轮没有任务相关增量时保留未知并返回调用者。阅读预算由调用者给出,不继承预测的页数和四维配额。

What it can do on your machine

Read from SKILL.md and the folder at commit 760d128. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships 2 files in scripts/ (Python and TypeScript), which the agent can run.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Common Retrieval loads about 399 tokens when it runs, and up to ~1.3k if it reads all its reference files. Until then it costs about 26 tokens; SKILL.md has 48 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~26
When it runs · the whole SKILL.md, loaded when a task matches
~399
With references · SKILL.md plus every file in references/, read only if the agent opens them
~1.3k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from lihaozheCharlie/the-way-here at commit 760d128, republished under its MIT licence (© lihaozheCharlie). 48 words, ~399 tokens.

Download SKILL.mdSave it as .claude/skills/common-retrieval/SKILL.md (or your agent's skills folder). This skill also uses 4 other files; get the full folder from GitHub.
name
common-retrieval
description
知识库共享检索:绑定资料范围,关键词召回、来源和时间辨析、渐进阅读、双链追溯与证据核验。由查询、预测、了解度和构建 Skill 按需调用;不负责回答格式、未来推演或写入。

通用:知识检索

查询、预测、评估或构建需要检索知识库时加载。调用者决定问题、检索深度、允许资料范围和停止标准;本 Skill 只提供检索与证据边界,不自动运行预测、不授权写入、不要求调用 consume/query。

绑定资料

  • 沿用任务选定的知识库 ID,从根配置读取 Wiki 与来源目录;整个检索过程不随 UI 切库。
  • 调用者提供冻结文件时,核对 ID 和 hash,只读该文件。禁止为了补线索转读实时库或其他库。扫描、预测各自保留允许的证据范围。
  • 普通查询和构建可从当前绑定库开始。简单事实查询可用现有只读工具直接查目标页,不强制全库快照。多轮检索需要稳定资料时按 工具说明 创建任务快照;它由当前资料生成,不修改原始文件。
  • 构建写入后旧快照不会自动更新。若需查询新内容,显式刷新同一个库的快照,区分前后版本,不把旧引用当新内容。

按问题逐步读取

  1. 先找最具体的页面或来源。综合问题通常从相关 Wiki 开始,原句核验从来源开始,构建从新增来源及对应目标页开始。索引用于定位,不必先读完整总入口。
  2. 用具体对象、标题、别名、关键词及近义表达召回,命中片段只用于选择阅读位置。标题/别名命中也可能相关;不要仅靠文件日期或词频决定重要性。
  3. 读相关段落及上下文;重要页面读全文。区分已发现的文件和实际读过的内容,截断片段不能冒充全文核验。
  4. 围绕缺口沿出链、反链追溯直接来源、同一事件、后续变化、修正和反例。每次扩展有一个具体目的,通常先一跳;泛主题链接和高连接度不自动表示重要。歧义/未解析链接只作为待核查线索。
  5. 对解释、趋势或决策所依赖的关键判断查反例和后续撤回;简单事实定位无需强制多轮。答案已获支持时停止,连续两轮没有任务相关增量时保留未知并返回调用者。阅读预算由调用者给出,不继承预测的页数和四维配额。

来源、时间和去重

工具按 来源提示配置 给目录角色和词语线索;这些不是语义分类结论。按段落判断本人经历、直接自述、他人经历/书摘、用户评论、愿望/计划、行动/结果和综合推断。书中第一人称可能属于作者;记过笔记只证明接触/关注,明确认同与实际行动要分别保留。索引、日志和生成回信不能充当独立个人事实,但查询这些内容本身时仍然可检索。

时间分记录日期、事件时间、适用范围和综合覆盖时间。源文件记录日期候选暂取 Start,其次文件名,冲突必须回读;正文日期、年份和相对时间按具体语句解释,不取全文最大日期。end 或文件修改时间不能证明当前状态。未来计划、历史回忆、条件假设不能混为已发生事件;未知保留范围或原表达。

同一事件的日记、综合、索引和回信合并为一条来源链。重要结论优先追原始证据;无法追溯则标记为综合判断。明确撤回限制旧判断,但较新的转述不自动推翻更直接的证据。记录频率不是生活占比,没搜到不等于从未发生。

交给调用者

交付相关页面、核实过的短引文和位置、来源角色/时间的不确定性、关联来源、重复与矛盾线索、仍未解决的缺口。无需统一报告模板或暴露内部推理。调用者负责回答、未来情景、了解度评分或 Wiki 修改;是否需要概率、四维覆盖和行动建议不在本层决定。

修改此 Skill、配置或工具后运行 common-quality-gate,并验证普通查询、冻结预测和构建检索均保持范围隔离。

© lihaozheCharlie, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 4 other files (scripts, references) in knowledge-engine/skills/common/retrieval of lihaozheCharlie/the-way-here.

  • SKILL.md
  • references/source-policy.json
  • references/tools.md
  • scripts/evidence_reader.py
  • scripts/snapshot.mts

Open the folder on GitHubat commit 760d128

Compare with similar skills

Common Retrieval next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Common Retrieval compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Common Retrieval this skilllihaozheCharlie/the-way-here145—~399Automated safety check: PassMIT
Iterative Retrievalaffaan-m/ECC277k7 repos~1.6kAutomated safety check: PassMIT
Iterative Retrievalaffaan-m/ECC277k2 repos~1.1kAutomated safety check: PassMIT
Iterative Retrievalaffaan-m/ECC277k2 repos~1.3kAutomated safety check: PassMIT
Iterative Retrievalaffaan-m/ECC276k1 repos~1.1kAutomated safety check: PassMIT
Retrieval Reflexgarrytan/gbrain31k—~735Automated safety check: PassMIT

Similar skills

  • Pattern for progressively refining context retrieval to solve the subagent context problem.

    277k GitHub starsUsed in 7 repos~1.6k tokens
    Agent WorkflowsAuto-check passed
  • サブエージェントのコンテキスト問題を解決するために、コンテキスト取得を段階的に洗練するパターン. An agent skill from affaan-m/ECC.

    277k GitHub starsUsed in 2 repos~1.1k tokens
    Auto-check passed
  • 서브에이전트 컨텍스트 문제를 해결하기 위한 점진적 컨텍스트 검색 개선 패턴. An agent skill from affaan-m/ECC.

    277k GitHub starsUsed in 2 repos~1.3k tokens
    Auto-check passed
  • 逐步优化上下文检索以解决子代理上下文问题的模式

    276k GitHub starsUsed in 1 repo~1.1k tokens
    Auto-check passed
  • Retrieval Reflex

    garrytan/gbrain

    When/what to retrieve — open the brain page for a salient entity before answering from memory.

    31k GitHub stars~735 tokensUpdated yesterday
    Auto-check passed
  • Remote Retrieval Automation

    ComposioHQ/awesome-claude-skills

    Automate Remote Retrieval tasks via Rube MCP (Composio). An agent skill from ComposioHQ/awesome-claude-skills.

    77k GitHub starsUsed in 3 repos~764 tokens
    Productivity & AutomationAuto-check passed

More from lihaozheCharlie/the-way-here

All 8 skills in this repo
  • Common Reasoning Lenses

    lihaozheCharlie/the-way-here

    为个人 Wiki 的查询、回信和解释性构建提供可动态发现的人物推理视角。它只改变怎样理解证据,不构成新证据,也不决定是否写入。

    145 GitHub stars~640 tokensUpdated yesterday
    Auto-check passed
  • Common Quality Gate

    lihaozheCharlie/the-way-here

    验证 wiki 或 Skill 修改是否满足证据、归档、前置元数据、生成索引安全、链接和完成标准;纯健康检查保持只读. An agent skill from lihaozheCharlie/the-way-here.

    145 GitHub stars~488 tokensUpdated yesterday
    Auto-check passed
  • Common Skill System

    lihaozheCharlie/the-way-here

    分析、设计、验证或更新本 wiki 的 Codex Skill 体系、AGENTS.md 唯一路由、职责图和验证规则。当前目标只要求分析时保持只读。

    145 GitHub stars~609 tokensUpdated yesterday
    Auto-check passed
  • Common Signal Detector

    lihaozheCharlie/the-way-here

    在普通对话、wiki 维护或路由诊断中扫描可沉淀信号;本 Skill 只分类,是否更新由 Agent 结合当前目标判断。

    145 GitHub stars~228 tokensUpdated yesterday
    Auto-check passed
  • Consume Query

    lihaozheCharlie/the-way-here

    以只读方式回答个人 wiki 问题:先检索综合页面,仅在需要证据时查原始笔记。不要用于摄取或修改 wiki. An agent skill from lihaozheCharlie/the-way-here.

    145 GitHub stars~391 tokensUpdated yesterday
    Auto-check passed
  • Consume Predict Self

    lihaozheCharlie/the-way-here

    预测页主动触发:只读绑定知识库的冻结资料,生成有证据的五年生活情景。

    145 GitHub stars~896 tokensUpdated yesterday
    Auto-check passed

Questions about Common Retrieval

What does Common Retrieval do?

知识库共享检索:绑定资料范围,关键词召回、来源和时间辨析、渐进阅读、双链追溯与证据核验。由查询、预测、了解度和构建 Skill 按需调用;不负责回答格式、未来推演或写入。. Common Retrieval is an agent skill from lihaozheCharlie/the-way-here.

How do I install Common Retrieval in Claude Code?

Run `npx skills add lihaozheCharlie/the-way-here --skill common-retrieval -a claude-code`. Or copy the skill folder (knowledge-engine/skills/common/retrieval in lihaozheCharlie/the-way-here) into .claude/skills/common-retrieval in your project. Claude Code loads it when a task matches its description.

How do I install Common Retrieval in Codex?

Run `npx skills add lihaozheCharlie/the-way-here --skill common-retrieval -a codex`. Or copy the skill folder (knowledge-engine/skills/common/retrieval in lihaozheCharlie/the-way-here) into .agents/skills/common-retrieval in your project. Codex loads it when a task matches its description.

Can I use Common Retrieval in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add lihaozheCharlie/the-way-here --skill common-retrieval -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/common-retrieval, .gemini/skills/common-retrieval, .github/skills/common-retrieval and .opencode/skills/common-retrieval in your project.

What does Common Retrieval need to run?

Going by SKILL.md and its folder, Common Retrieval needs Python and TypeScript for the scripts in its folder. Our summary lists: Python 3.

Does Common Retrieval access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Common Retrieval safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does Common Retrieval use?

Common Retrieval is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Common Retrieval use?

About 399 tokens (SKILL.md is roughly 1.6k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 932 tokens, read only when the agent opens those files.

What are the alternatives to Common Retrieval?

Skills that share tags, products or a category with Common Retrieval: Iterative Retrieval (affaan-m/ECC, 277k stars), Iterative Retrieval (affaan-m/ECC, 277k stars), Iterative Retrieval (affaan-m/ECC, 277k stars) and Iterative Retrieval (affaan-m/ECC, 276k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Common Retrieval?

lihaozheCharlie (a GitHub user) maintains it in lihaozheCharlie/the-way-here, which has 145 GitHub stars. The repository holds 8 skills in this directory. The repository was last updated on October 9, 2026.

Source: lihaozheCharlie/the-way-here on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.