agent-md-refactor
Refactor bloated AGENTS.md, CLAUDE.md, or similar agent instruction files to follow…
Navigate difficult conversations and deliver constructive feedback using structured frameworks. Covers the Preparation-Delivery-Follow-up model and Situation-Behavior-Impact (SBI) feedback technique. Use when preparing for difficult conversations, giving feedback, or managing
$ npx -y skills add softaworks/agent-toolkit --skill feedback-mastery --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/feedback-masteryContext preview
The summary Claude sees to decide when to auto-load this skill.
Navigate difficult conversations and deliver constructive feedback using structured frameworks. Covers the Preparation-Delivery-Follow-up model and Situation-Behavior-Impact (SBI) feedback technique. Use when preparing for difficult conversations, giving feedback, or managing
name: feedback-mastery description: Navigate difficult conversations and deliver constructive feedback using structured frameworks. Covers the Preparation-Delivery-Follow-up model and Situation-Behavior-Impact (SBI) feedback technique. Use when preparing for difficult conversations, giving feedback, or managing conflicts. allowed-tools: Read, Glob, Grep
This skill provides frameworks for navigating difficult workplace conversations and delivering effective feedback. Whether you're addressing performance issues, resolving conflicts, or giving constructive feedback, these structured approaches lead to better outcomes.
**Core insight:** Research shows that employees who approach difficult conversations with preparation and a clear framework are **60% more likely to reach a positive resolution** than those who engage without a plan.
Use this skill when:
**Keywords**: feedback, difficult conversation, 1:1, one-on-one, performance, conflict, expectations, behavior, confrontation
A three-part structure for difficult conversations:
| Phase | Focus | Key Questions | | --- | --- | --- | | **Preparation** | Understand the issue, define goals, manage emotions | What's the problem? What outcome do I want? Am I calm? | | **Delivery** | Open neutrally, use facts not blame, encourage dialogue | How do I start? What evidence do I have? How do I involve them? | | **Follow-up** | Document actions, set check-ins, provide support | What did we agree to? When will we check in? How do I support? |
**Situation-Behavior-Impact (SBI)** structures feedback to be specific, objective, and actionable:
| Component | Description | Example | | --- | --- | --- | | **Situation** | Describe the specific context | "During yesterday's code review..." | | **Behavior** | State the observable action (not interpretation) | "...you interrupted Sarah three times while she was explaining her approach..." | | **Impact** | Explain the effect on team/project/person | "...which made her hesitate to share ideas and slowed down our discussion." |
**Why it works:** SBI removes assumptions and focuses on observable facts, reducing defensiveness.
Ask yourself:
Before the conversation, clarify what you're seeking:
| Goal Type | Example | | --- | --- | | Behavior change | "I want them to submit code reviews on time" | | Mutual understanding | "I want to understand what's blocking them" | | Expectation setting | "I want to clarify what 'done' means for features" | | Problem solving | "I want to find a solution together" |
**Tip:** Use if-then statements to clarify stakes: > "If this behavior continues, then the project timeline will suffer, leading to missed deliverables."
High emotional intensity reduces cognitive processing by 30%. Before the conversation:
**Reframing technique:**
| Accusatory | Constructive | | --- | --- | | "You always miss deadlines and it slows everyone down" | "I've noticed some recent delays and want to understand any challenges you're facing" | | "You never test your code properly" | "I've seen a few bugs slip through recently. Let's talk about our testing process" |
1. **Open with neutrality and intent** 2. **Present the issue using facts, not blame** 3. **Encourage dialogue and solutions**
| Context | Opening | | --- | --- | | General | "I want to talk about something important to our team's success, and I'd love to hear your perspective." | | Performance | "I've noticed some patterns I'd like to discuss. My goal is to support you, not criticize." | | Conflict | "I sense there might be some tension, and I'd like to understand what's happening from your side." | | Expectations | "I want to make sure we're aligned on expectations. Can we talk through how this project is going?" |
| Blaming | Factual | | --- | --- | | "You're not committed to this project" | "I've noticed your updates have been brief in our last three meetings. Is something affecting your workload?" | | "You don't care about code quality" | "This PR had 12 bugs caught in QA. Let's talk about what happened and how we can improve" | | "You're always late" | "The standup started at 9:00 and you joined at 9:15 the last three days. What's going on?" |
**Key principles:**
After stating your observation, shift to collaboration:
| Situation | Dialogue Prompt | | --- | --- | | Understanding barriers | "What's been challenging about this?" | | Seeking their view | "How do you see the situation?" | | Finding solutions | "What would help you succeed here?" | | Checking alignment | "Does this match your understanding of what happened?" |
Even successful conversations need follow-through to create lasting change.
Opinionated skills shared by @leonardocouy for improving daily work efficiency with Claude Code. Skills are packaged instructions and scripts that extend agent capabilities across development, documentation, planning, and professional workflows.
Repo: softaworks/agent-toolkit
Refactor bloated AGENTS.md, CLAUDE.md, or similar agent instruction files to follow…
Create API handoff documentation for frontend developers. Use when backend work is complete…
Generate architecture documentation using C4 model Mermaid diagrams. Use when asked to create…
Use when the user asks to run Codex CLI (codex exec, codex resume) or references OpenAI Codex…
This skill should be used when creating a Claude Code slash command. Use when users ask to…
Create high-quality git commits: review/stage intended changes, split into logical commits,…