# Survey Peers

> Map what else fills the same role at the same level of detail, and where the gaps between them lie

- Skill: `synapseradio/survey-peers` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx get-skill add synapseradio/survey-peers`
- Raw SKILL.md: https://api.skillmd.com/api/skills/synapseradio/survey-peers/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: synapseradio (https://skillmd.com/u/synapseradio)
- Updated: 2026-11-15
- Page: https://skillmd.com/skills/synapseradio/survey-peers

---


## Steps

### 1. What role does the subject fill, and at what grain?

Name the question it answers or the job it does, and the level of detail it sits at.

### 2. What else fills that role at that grain?

List the peers, including ones from outside the current domain. Keep every peer at the subject's grain: no parts of it, no systems containing it, no classes above it.

### 3. What most separates each peer from the subject?

Name one property per peer.

### 4. How do the peers arrange?

Place them along the one or two differences that separate them most.

### 5. Where are the empty positions?

Name each position in the arrangement that nothing occupies, and say whether it is empty by accident or for a reason.

### 6. Where does the subject sit?

Report its position among the peers and its nearest neighbours.

## Output

Choose the mode from what you were handed:

- A question set: answer it.
- A request for questions only: write the question set and stop.
- Anything else: write the question set, then answer it.

Write the question set in this form: one question per step, in step order, each reworded to name the subject, so that a reader holding only the set can answer it.

```markdown
**Subject:** [what is examined, named so it can be found without this conversation]

**Given:**

- [a fact, source, or limit the answers must respect] ([where it came from])

**Questions:**

1. [step 1's question, naming the subject]
2. [step 2's question, naming the subject]
3. [step 3's question, naming the subject]
```

Answer in this form, keeping each question's number and wording, from the set and from what you can check yourself:

```markdown
**Subject:** [the subject, as the set names it]

1. [step 1's question, naming the subject]

   [the answer]

2. [step 2's question, naming the subject]

   Skipped: [why this step does not apply]

3. [step 3's question, naming the subject]

   Open: [the fact the answer needs that the set does not give]
```

