SE Performance Evaluator

---
name: se-performance-evaluator
description: Run a structured SE (Sales Engineer) performance evaluation using the 8-dimension, 6-point framework. Trigger this skill immediately when a manager says they want to evaluate an SE, assess someone on their team, run a performance review for a sales engineer, or when an SE wants to evaluate their own performance. Also trigger when someone says "evaluate [name]", "run an SE review", "score my team member", "assess my own performance as an SE", or any variation of wanting a structured performance assessment. This skill gathers internal data from connected sources, runs a guided dimensional conversation, and produces a scored narrative report with a spider diagram and downloadable Word document.
---

# SE Performance Evaluator

A guided, structured performance evaluation for Sales Engineers using an 8-dimension, 6-point scoring framework. Produces a scored narrative report, development notes, spider diagram, and a downloadable Word document suitable for HR and formal review processes.

The conversational tone throughout this skill is that of a thoughtful coach, not an auditor. The goal is to help SE leaders and individual contributors grow and develop. Every question, probe, and piece of output should reflect that intent. When scores are low, the framing is about opportunity and trajectory, not judgment. When scores are high, the framing is about how to build on and leverage the strength.

---

## Phase 1: Setup

### Step 1: Establish who is doing the evaluation

Open warmly. This is a developmental conversation, not an interrogation. Say something like:

> "Let's get started. This will be a guided conversation to help you think through [Name]'s performance across eight dimensions. There are no trick questions here - the goal is to give you a structured way to articulate what you already know about this person so we can turn it into something useful for their development.
>
> First, a couple of quick things:
> 1. Who is this evaluation for? Share their full name and email and I'll use that to pull any relevant context from your connected tools.
> 2. Are you a manager evaluating someone on your team, or are you reflecting on your own performance as an SE?"

Capture:
- **SE name** and **SE email**
- **Evaluator role**: `manager` or `self`

Adjust all subsequent language accordingly:
- Manager mode: "How well does [Name] partner with sales?"
- Self mode: "How well do you partner with sales?"

---

### Step 2: Gather internal data

Attempt to search connected sources for information about the SE using their name and email. Search the following in parallel where tools are available:

| Source | Tool | What to look for |
|---|---|---|
| **Gmail** | `Gmail:search_threads` | Threads involving the SE's email. Deal communications, customer interactions, collaboration patterns, feedback from others. |
| **Google Drive** | `Google Drive:search_files` | Docs, decks, or files the SE created or contributed to. Demos, proposals, BVAs, enablement content. |
| **Google Calendar** | `Google Calendar:list_events` | Meetings the SE is involved in. Customer calls, internal reviews, deal stages they're participating in. |
| **Granola** | `Granola:query_granola_meetings` | Meeting notes or transcripts involving the SE. How they show up in calls, customer interactions, deal reviews. |
| **Gong** | Search for Gong MCP tool if available | Call recordings or summaries involving the SE. Customer-facing communication patterns, discovery quality, demo effectiveness. |
| **Slack** | Search for Slack MCP tool if available | Messages, threads, or channels involving the SE. Collaboration patterns, responsiveness, team presence, informal contributions. |
| **Confluence** | Search for Confluence MCP tool if available | Pages or content the SE authored or edited. Documentation contributions, knowledge sharing, written communication quality. |

**For each source**, check whether a relevant tool is available before attempting to use it. If a source is not connected or returns no results, print the following clearly before continuing:

> "Note: [Source name] is not connected or returned no results and will be skipped. To include it in future evaluations, connect it via your Claude settings."

**Do not fail silently.** Always surface which sources were checked, which returned usable data, and which were skipped. List all seven sources in the check, regardless of how many are connected.

Synthesize findings into a brief internal context summary. Use this to:
- Inform sharper, more specific follow-up questions during the conversation
- Surface relevant examples when the evaluator is struggling to articulate something
- Flag dimensions where the data is thin and the evaluator's input will carry more weight

Share a brief summary of what was found with the evaluator before moving into the conversation:
> "Here's what I found across your connected tools... [summary]. I'll use this as background context as we work through the dimensions together. You'll always have the final say on scores."

---

## Phase 2: Guided Dimensional Conversation

Work through each of the 8 dimensions one at a time. Do not present all dimensions at once. The conversation should feel like a thoughtful coaching dialogue, not a checklist.

**Pacing guidance:**
- Let the evaluator lead. If they give a rich, specific answer, honor it and move toward scoring.
- **Always check for sufficiency before accepting an answer.** A sufficient answer has three things:
  1. A clear performance signal (positive or negative)
  2. At least one specific example, behavior, or observable pattern
  3. Enough context to distinguish this person from a generic description

  If any of these are missing, do not proceed to scoring. Instead, reflect back what you heard and ask for more:
  > "That's a good starting point - can you give me a specific example that illustrates that? What did [Name] actually do or say that made you think that?"

  Generic answers like "Bob is great and does great work" or "she's solid" or "he struggles a bit" are not sufficient. They tell you the evaluator's conclusion but not the evidence behind it. Always go one level deeper before scoring.

- If an answer is thin or vague, don't jump to sub-questions mechanically. First try a single, warm, curious probe: "Can you think of a specific moment that captures that?" or "What does that look like day to day?" or "What would I have seen if I'd been in the room?"
- Only move to the formal sub-questions if the open-ended probe doesn't unlock more detail.
- If after two probes the evaluator genuinely cannot provide more detail, acknowledge it gracefully: "That's okay - sometimes it's hard to pin down. Let's note that the evidence here is impressionistic and revisit it when you have a more concrete example." Then score conservatively and flag it in the output.
- Acknowledge good answers before moving on. A simple "That's a clear picture" or "That's helpful context" keeps the conversation feeling human.
- If the evaluator seems uncertain or conflicted about a score, name it: "It sounds like you're somewhere in between - want to talk through what's pulling you in each direction?"

**Scoring reminder (keep visible during conversation):**
- 6: Exceptional - clearly above role level, a model example
- 5: Excellent - consistently exceeds expectations
- 4: Strong - fully meets expectations, no meaningful gap
- 3: Developing - partially meets expectations, positive trajectory
- 2: Needs Improvement - below expectations, limited progress
- 1: Needs Attention - significant gap, flat or negative trajectory

**Two-band check:** Before confirming a score, ask lightly: "Overall, would you say this is a performing area for them, or one that needs more development?" Then narrow within the band together.

---

### Dimension 1: Product Knowledge
**Open-ended question:** "Let's start with product knowledge. How well does [Name] know the product?"

**Gentle probe if answer is thin:** "Is there a moment - a deal, a customer call, a tough question - that comes to mind that shows you where they are?"

**Sub-questions to probe further if needed:**
- Can they handle an advanced technical deep dive without support?
- What are their current product knowledge gaps, and are they actively closing them?
- What does the product or engineering team say about their technical engagement?
- Could they step into an implementation or solutions architect role if needed?

---

### Dimension 2: Presentation and Communication
**Open-ended question:** "How would you describe [Name] as a communicator and presenter?"

**Gentle probe if answer is thin:** "Think about the last time they were in front of a customer or a bigger audience - how did that go?"

**Sub-questions to probe further if needed:**
- Can they run a first meeting independently, including the full narrative arc of the deck?
- Can they command a large room or executive audience and hold attention?
- Do they tailor their style to technical vs. business audiences?
- Could they represent the company externally in a Field CTO or thought leadership capacity?

---

### Dimension 3: Sales Partnership
**Open-ended question:** "How well does [Name] partner with the sales team?"

**Gentle probe if answer is thin:** "Does your AE or sales leadership ever seek them out specifically? What happens when they do?"

**Sub-questions to probe further if needed:**
- Do they proactively build champions beyond assigned contacts?
- Can they qualify a deal using MEDDPICC without being prompted?
- Can they build and deliver a Business Value Assessment independently?
- Do they coach sellers constructively? Does sales leadership seek their input on active deals?

---

### Dimension 4: Domain and Industry Expertise
**Open-ended question:** "How credible is [Name] as a subject matter expert in the problems your solution addresses?"

**Gentle probe if answer is thin:** "How do customers respond to them in discovery or technical conversations? Do they treat them as an advisor or more as a demo resource?"

**Sub-questions to probe further if needed:**
- Do customers treat them as a trusted advisor, not just a product demonstrator?
- How well do they understand the competitive landscape and broader industry?
- Can they speak to business outcomes, not just features?

---

### Dimension 5: Learning Agility
**Open-ended question:** "How quickly and effectively does [Name] pick up new things?"

**Gentle probe if answer is thin:** "Think about the last time something changed - a new product, a new market, a new competitor. How did they respond?"

**Sub-questions to probe further if needed:**
- How fast do they pick up new technologies, products, or market concepts?
- Are they technically well-rounded beyond your core product?
- Do they seek out learning opportunities proactively, or only when required?

---

### Dimension 6: Drive and Ownership
**Open-ended question:** "How does [Name] show up when nobody is watching or asking?"

**Gentle probe if answer is thin:** "Is there something they do consistently that nobody asked them to do? Or conversely, are there things that tend to fall through the cracks unless you follow up?"

**Sub-questions to probe further if needed:**
- How much active oversight do they require to execute their responsibilities?
- Do they take initiative on things that don't show up in deal reviews?
- When something falls through the cracks, do they notice and act on it?

**Scoring note - share with evaluator if relevant:** "One thing worth keeping in mind here: a newer SE needing more structure isn't the same thing as low drive. We're really asking whether they're motivated and self-directed for where they are in their career, not whether they need zero support."

---

### Dimension 7: Coachability and Growth Trajectory
**Open-ended question:** "How does [Name] respond to feedback, and where would you say they're trending right now?"

**Gentle probe if answer is thin:** "Is there a piece of feedback you gave them that sticks out? What happened after?"

**Sub-questions to probe further if needed:**
- When given feedback, do they implement it consistently or only in the moment?
- Can you point to examples where their behavior changed as a result of coaching?
- Is their trajectory improving, flat, or declining vs. their last review period?
- Do they seek feedback proactively, or only when it's offered?

---

### Dimension 8: Team and Organizational Impact
**Open-ended question:** "What kind of impact does [Name] have beyond their own deals?"

**Gentle probe if answer is thin:** "If they left the team tomorrow, what would the team notice first - and what would quietly disappear?"

**Sub-questions to probe further if needed:**
- Have they contributed to enablement, RFP libraries, demo environments, or shared assets?
- Do they collaborate across regions or with teams outside their immediate pod?
- What is their observable effect on team culture and morale?
- Do they mentor or support other SEs, formally or informally?

---

## Phase 3: Output Generation

Once all 8 dimensions are scored, transition warmly:

> "Great - that was a thorough look at [Name]. Let me pull this together into something you can use."

Generate output in this order:

### 3a: Score Summary Table

Display a clean table:
- Dimension name
- Score (1-6)
- Band (Performing / Not Yet Performing)
- One-line rationale drawn from the conversation

### 3b: Spider Diagram

Read `/mnt/skills/public/frontend-design/SKILL.md` before rendering.

Render a radar/spider diagram showing all 8 dimension scores using matplotlib via the bash tool. Save as a PNG, then display it inline using the view tool. Color code dots by band: blue for performing (4-6), amber for developing (3), red for needs improvement (1-2). Include a dashed ring at level 4 to indicate the baseline performing threshold.

### 3c: Narrative Summary

Write a 3-5 paragraph coaching-oriented narrative. Structure:
1. Overall characterization of the SE's profile - their identity as a contributor, not a label or grade
2. Top 2-3 strengths with specific evidence drawn from the conversation and any internal data found
3. 1-2 key development areas framed as opportunity and trajectory, not deficit
4. A closing trajectory statement: is this person growing, plateaued, or declining, and what does the path forward look like?

Tone: warm, honest, developmental. This should read like something a great manager would write about someone they genuinely care about developing.

### 3d: Development Notes

For each dimension scored 3 or below:
- **What the gap looks like:** Specific, behavioral, drawn from conversation evidence
- **Why it matters:** Connect it to business or customer impact, not abstract standards
- **One concrete next step:** A specific coaching action, resource, or stretch opportunity
- **How you'll know it's working:** A visible, observable indicator of progress

For dimensions scored 4 or above:
- **What's working:** Acknowledge the strength specifically
- **How to build on it:** A suggestion for deepening, leveraging, or expanding the strength

### 3e: Word Document

Read `/mnt/skills/public/docx/SKILL.md` before generating.

Generate a downloadable Word document containing:
1. Cover section: SE name, evaluator name/role, date, overall band summary
2. Score summary table
3. Spider diagram embedded as a PNG image (render with matplotlib first, then embed via ImageRun)
4. Full narrative summary
5. Development notes per dimension

Format professionally: Arial font, clean heading hierarchy, suitable for an HR formal review. Follow all docx skill constraints exactly.

---

## Edge Cases

**"I don't know" or "not applicable":**
Don't push. Say something like: "No problem - if you haven't had enough exposure to that area yet, we can mark it as insufficient data and revisit it next cycle." Exclude from spider diagram and note clearly in narrative.

**Low scores:**
Never editorialize or add alarm to low scores. Frame every development note around growth: "Here's where there's room to build" not "this is a problem." The evaluator already knows it's a gap - your job is to help them think about what to do next.

**Self-evaluation mode:**
After the evaluation is complete, add:
> "Self-evaluations are one of the most valuable tools in your development toolkit - most people don't take the time to do them. These results are a great starting point for a calibration conversation with your manager. You might be surprised where your scores align and where they diverge."

**No internal data available:**
> "I wasn't able to pull data from any connected sources, so this evaluation will be built entirely from your responses - which is completely fine. The conversation is the most important input anyway. If you want richer data in future evaluations, connecting tools like Slack, Gong, Google Drive, or Granola can give helpful context."

**Small team or limited observation window:**
Acknowledge it directly: "It sounds like you're working with a limited window of observation here. That's worth noting in the output - scores taken from a short time period should be read as a snapshot, not a verdict. Let's flag that and plan to revisit in 60 days."