Skip to content
Agentmart
Submit a tool

xy

Resolve XY problems — surface questions that mask a deeper goal.

Skill by EATSTEAK

IndexedMIT
Write a review

Chalk Labs

Need help deploying xy?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/eatsteak-xy

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Resolve XY problems — surface questions that mask a deeper goal. Use when the user invokes /xy, when a request is vague or mismatched to its context, or when a narrow how-to looks like it's hiding the real objective.

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/xy

What's inside

Skill name
xy
Description
Resolve XY problems — surface questions that mask a deeper goal. Use when the user invokes /xy, when a request is vague or mismatched to its context, or when a narrow how-to looks like it's hiding the real objective.
License
MIT
Instructions
60 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Resolve XY problems

A request often names a means (Y) while the user's real goal (X) stays unsaid. Don't answer Y until you've checked it's the actual problem. Ask **one question at a time**; if an answer lives in the codebase, explore it (Grep/Read/Glob) instead of asking.

## Step 1 — Name Y and its assumptions

Restate the request in one line as Y, then list the assumptions baked into it. Don't judge yet — just surface them.

If Y is a plain fact question with no deeper goal behind it (a dictionary lookup, a translation, a syntax answer), say so, answer Y, and stop here.

**Done when:** Y is stated and every assumption behind it is written down.

## Step 2 — Walk toward X

Ask one question at a time from the list below, following whichever branch the answer opens. Aim to find X — what they're ultimately trying to achieve.

- **Ultimate goal:** "What are you ultimately trying to do?"
- **Source of Y:** "Where did this approach come from — an error, a tutorial, a guess?"
- **equivalence:** "If I fully solved Y right now, would your real goal be met?" → "no / not sure" is the strong XY signal.
- **Constraints:** "What's forcing Y — a deadline, a tool, a policy, a guess at what's possible?"
- **Prior attempts:** "What have you tried, and what happened?"

**Done when:** you can state X in one sentence, and you've run the equivalence check.

## Step 3 — Diagnose

Pick one verdict:

- **XY confirmed** — Y won't reach X even if solved.
- **Y is legit** — Y is actually the best path. Answer it and stop.
- **Underspecified** — still can't tell; return to Step 2.

XY signals: Y is unusually specific or rare; X is never mentioned; solving Y ≠ reaching the goal; Y copies someone else's fix verbatim.

**Done when:** one verdict is chosen, with a one-sentence reason.

## Step 4 — Reframe and propose (then stop)

Output this block:

```markdown
## Reframed problem
- Surface question (Y): ...
- Real goal (X): ...
- Why Y was a detour: ...

## Recommended approach
1. [best path to X directly — one line + why]
2. [alternative, with its trade-off]

## Next step (your call)
- A) ...  ← recommended
- B) ...
```

If Step 3 said "Y is legit", say so and answer Y briefly instead.

**Done when:** the block is written. Do **not** start implementing — hand the choice back to the user.

From EATSTEAK/skills, MIT.

More from EATSTEAK/skills

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
Jul 1, 2026 (4 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from EATSTEAK

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get xy running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying xy?

CallGet help
Cite this page

xy skill. Agent Mart. https://agentmart.dev/skills/eatsteak-xy (last updated 2026-10-24)