Token导航 LogoToken导航

Agent Skill · 研究检索

feedback(feedback 数据库)

feedback 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

feedback 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为只读;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 52,870。

总安装 544周安装 22GitHub Stars 21下载 171核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:feedback(feedback 数据库)
来源仓库:https://github.com/manager-dot-dev/manager-skills
仓库路径:skills/feedback
安装命令:
npx skills add https://github.com/manager-dot-dev/manager-skills --skill feedback
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/manager-dot-dev/manager-skills --skill feedback

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

feedback 用于查找、检索和筛选相关信息。适合在 Codex、Claude、Cursor、Gemini CLI 中根据关键词或任务场景快速定位候选结果。通过 npx skills add 命令从指定 GitHub 仓库安装使用。需确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写操作。

建议结合原始 README 核验具体用法后再部署到生产环境。

来源仓库内路径:skills/feedback

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/manager-dot-dev/manager-skills --skill feedbac…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

feedback 怎么安装?

在终端执行 npx skills add https://github.com/manager-dot-dev/manager-skills --skill feedback。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

feedback 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

feedback 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

feedback 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

feedback 用的人多吗?

公开数据是累计安装 544 次,近一周安装 22 次,下载 171 次,GitHub 21 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

feedback 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 只想先试试效果、不希望它改动本地文件的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
feedback

feedback 数据库 · 本页

54421只读未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Feedback

Before Starting

Check for EM context first:

  1. Read .agents/em-context.md if it exists
  2. If a person is mentioned, look for .agents/reports/[name].md — it may contain their feedback preferences, prior feedback history, and working style
  3. Use that context — only ask for information not already covered

If .agents/em-context.md does not exist, ask for a minimal manager profile first and save it before giving detailed advice: role/title, team size, team mission or ownership area, and current challenge or priority.

If a specific person is central to the conversation and .agents/reports/[name].md does not exist, ask for a minimal profile for that person first and save it before giving detailed advice: title/level, tenure, strengths, and current challenge or growth area.

If the conversation reveals durable new context later, update .agents/em-context.md or .agents/reports/[name].md automatically. Save stable facts and patterns, not guesses, transient frustration, or unresolved interpretations.

Response Style

Keep the first answer concise and useful. Do not dump the whole framework unless the user asks for depth.

Default to:

  • State the likely diagnosis or recommendation first
  • Ask at most 2-3 targeted questions only if the missing context changes the advice
  • Give the next concrete action and, when useful, exact wording the manager can use
  • Mention the relevant framework briefly, but do not explain every part of it
  • Offer a deeper version only after the direct answer

How to Use This Skill

Identify what the user needs before diving into frameworks:

  • Giving positive feedback → Go to Precision: No Weasel Words — it applies equally to praise
  • Constructive feedback — know the issue but not how to open → Opening a Critical Feedback Conversation, then Instead of the Compliment Sandwich
  • Expecting resistance or pushback → Behavior Change Stages first
  • Something happened recently and the user wants to address it now → In-the-Moment Feedback (don't save it for the 1:1)
  • Building feedback culture across the whole team → Speed Feedback
  • Writing feedback for a review or document → Precision section first; then ask for the specific events to document
  • Building a lightweight weekly accountability loop with reports → read references/extended.md — Weekly Called Shots
  • Getting honest feedback from your own team → Getting Feedback from Your Team

If the user hasn't given you a specific event yet, ask for it. Vague feedback inputs always produce vague feedback outputs.


Default Response Shape

When helping with feedback, produce wording the manager can actually say:

  1. Intent: what outcome the feedback should create.
  2. SBI draft: situation, behavior, impact, in concrete language.
  3. Opening line: a direct but humane start to the conversation.
  4. Follow-up question: invite their perspective without softening the message.
  5. Next step: request, agreement, or observation period.

For praise, keep it specific. For corrective feedback, do not hide the message inside a compliment sandwich.


Precision: No Weasel Words

The most common feedback mistake is vagueness. "You're too blunt in meetings" or "your communication needs improvement" — these feel like feedback but give the person nothing to act on. They're also easy to dismiss: "I don't think I'm blunt."

Specific feedback is factual and indisputable.

Weak: "You're too blunt in meetings." Specific: "In the meeting with Christopher, you cut him off when he started suggesting the React-based solution. We missed an opportunity to check more approaches. I also noticed that Christopher stopped contributing to other meetings after that."

The specific version names the event, the person affected, and the downstream consequence. The other person can't dispute the facts — they can only engage with what happened.

Apply this same standard to positive feedback. "You did a great job" is forgettable. "The way you handled the incident on Thursday — staying calm, communicating clearly to stakeholders while the team debugged — that's exactly what we need in situations like that" is memorable and repeatable.

Before giving feedback, do the homework. Identify the specific event, the specific behavior, and the specific impact. Vague feedback isn't kind — it's just unprepared.


Public Praise Isn't Universal

"Praise in public, criticize in private" is standard management advice. The first half isn't always true.

Some engineers genuinely don't want to be called out in front of the team. They prefer private acknowledgment. Surprising them with public praise — a Slack shoutout, a nomination for a recognition program — can create awkwardness instead of motivation.

Ask before you assume. A simple "How do you prefer to be recognized when you do great work?" in a 1:1 is enough.


Translating Social Cues

Not everyone can read a room. If you're good at this, part of your job is translating for those who aren't.

When a team member says something that lands badly and doesn't notice — upset someone, came across as dismissive, answered too abruptly in a meeting — tell them privately. Not as criticism, but as information: "I think that landed differently than you intended. Here's what I noticed." They often don't know. Telling them is the most useful thing you can do.


Behavior Change Stages

When addressing a performance or behavior problem, people typically move through three stages before accepting feedback:

  1. Ignore — they don't acknowledge the problem exists
  2. Deny — they actively push back ("That's not true" / "You're wrong")
  3. Blame others — they redirect responsibility elsewhere

In a single 1:1, you can often move someone through all three stages. The right response at each stage is different:

  • At Ignore/Deny: bring specific facts. Don't argue — present evidence calmly.
  • At Blame: don't let them escape into it. Acknowledge partial truths, then redirect back to what they can control.

The goal isn't to win the argument — it's to get to a point where the person can own the problem.


Instead of the Compliment Sandwich

Starting with a compliment and ending with one — sandwiching criticism in between — backfires in two ways: the opening compliment feels insincere (they're braced for the hit), and the closing compliment can bury the critique due to recency effects.

Four steps that work better (from Adam Grant, based on research):

  1. Explain why you're giving the feedback. Signal that you're on their side: *"I'm giving you this because I have high expectations and I know you can reach them."* It's harder to reject a hard truth from someone who believes in you.
  2. Take yourself off a pedestal. Reduce the power asymmetry: *"I've benefited from people giving me tough feedback — I'm trying to pay that forward."* It makes the feedback less threatening.
  3. Give the criticism directly. No burying. Say the specific thing, the specific event, the specific impact.
  4. Invite them to respond. Ask what they think. Let them engage, not just receive.

Opening a Critical Feedback Conversation

From *The Making of a Manager* (Zhuo): the moments right before a difficult feedback conversation feel uncomfortable — for both people. That discomfort is a good sign. It means the conversation is real.

Comfortable feedback conversations often aren't achieving anything. If neither person feels any friction, the feedback is probably not specific enough, not honest enough, or not about anything that actually matters.

What to say at the start:

The opening doesn't need to be elaborate. What it needs to signal is: you're saying this because you care about the person, not because you're judging them.

*"I want to share something that's hard for me to say, and I'm saying it because I think it matters for where you're going."*

Or simply naming what's happening: *"I have some direct feedback for you. It might feel uncomfortable, and I want you to know it's coming from a place of caring about your success here."*

Then say the thing. Don't circle around it. The opening sets the tone; the feedback itself has to be clear, specific, and direct. Softening after a strong opener defeats the purpose.

Silence is data. If someone goes quiet after critical feedback, don't rush to fill the space. Let them process. The silence often means they're actually taking it in — which is exactly what you wanted.


In-the-Moment Feedback

Waiting for the next 1:1 to give feedback costs most of its effectiveness. The closer feedback is to the event, the more likely the person can connect it to what they actually did.

Standup feedback — feedback given in the flow of work, at or near the moment it's relevant. Not a big formal conversation; a direct, brief comment. Example: a developer repeatedly delays code reviews for teammates. Raising this in the standup (or right after, in the hallway) is far more effective than bringing it up three days later in a 1:1 where the person may barely remember the specific instance.

The volume button. In-the-moment feedback doesn't need to be loud or intense. Think of it as tuning down the tone and volume — not the substance. Keep the content clear and direct; deliver it calmly and without drama. Brief, respectful, and close to the event beats formal, thorough, and a week later.

Why timing matters. Humans (like any animal) struggle to connect feedback to action when the gap is too long. Waiting days doesn't just reduce effectiveness — it trains the team to expect feedback only in formal settings, which reduces psychological safety in day-to-day work. Google's research on team effectiveness identified psychological safety as the #1 factor — and the team leader is the primary influence on it.


Speed Feedback: Building a Feedback Habit on the Team

Teams that only give feedback in formal reviews are too slow. Speed feedback is a structured activity for building a team-wide feedback habit.

How it works: Team members pair up in a private space. Each person gets a fixed amount of time to give feedback to the other, then they switch. Pairs rotate until everyone has spoken with everyone else.

The goal isn't depth — it's normalization. Early rounds focus on appreciation and celebrating contributions. As the team gets comfortable, they naturally start including improvement suggestions. The activity makes feedback part of the team's rhythm rather than a once-a-year event.

Works best with 5–8 people. Can be run remotely with breakout rooms and a shared board. Does not replace regular 1:1 feedback — it supplements it.


Getting Feedback from Your Team

Bad managers have many problems, but most of them know it. Good managers have a few invisible ones. When you ask for feedback in a performance review, you'll usually get silence — or "everything's great!" There is always something you can improve.

Three methods that get past the surface:

1. Try different angles. Don't accept the first "I can't think of anything." Keep asking from a different direction: "A behavior you didn't like?" "Something specific to your growth or our 1:1s?" "Something the team as a whole could have done better?" Real feedback is often only reachable because the question kept changing.

2. Share your own difficulties. When you share what you struggle with, people become more comfortable criticizing you. It signals that honest assessment is welcome. If your team only finds out about your struggles via a LinkedIn post, the psychological space for that conversation doesn't exist in your 1:1s.

3. Make it about a specific situation. People won't easily criticize you as a person — but they will talk about a project or an event. Instead of "where can I improve?" ask "what could we have done better in project X?" Then push further: "if we had to do the same project again, what would you change?" The second question is where the real feedback lives.


Dive Deeper

If the user asks where a framework came from, wants to read the original article, or wants more context on any topic in this skill — read references/sources.md for the full list of source articles (with links), books, and course material.


Related Skills

  • 1on1s — Delivering feedback in a 1:1 context
  • performance-reviews — Formal written feedback at review time

平台分布

Codex

35.81%
按下载量换算61

Claude

33.31%
按下载量换算57

Cursor

18.09%
按下载量换算31

Gemini CLI

10.43%
按下载量换算18

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills