Skip to content
Agentmart
Submit a tool

distill

/distill — Use when compressing a long document, transcript, or codebase into its essence.

Skill by kronael

IndexedUnlicense★ 6
Write a review

Chalk Labs

Need help deploying distill?

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/kronael-distill

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

/distill — Use when compressing a long document, transcript, or codebase into its essence. NOT for skill creation (use learn) or recall (use recall-memories).

Use cases: documents

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 and code. 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/distill

What's inside

Skill name
distill
Description
/distill — Use when compressing a long document, transcript, or codebase into its essence. NOT for skill creation (use learn) or recall (use recall-memories).
License
Unlicense
Instructions
109 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Distill — the 5/3 approach

Compression reveals truth: noise falls away, essence emerges. All state lives
in `.distill/` under the project root, so a run resumes from wherever it
stopped.

## Where it runs

A distillation reads a whole codebase or research set. Invoked in the main
thread, ALWAYS select a fact-gathering worker through `dispatch`, with the
subject, scope and required output; the worker loads this skill and presents
`.distill/final.md` when it returns. A subagent that loaded this skill runs the
phases itself.

## Start here — read the state, pick the phase

| `.distill/` holds | Phase |
|---|---|
| nothing | 1 — decompose |
| `areas.md`, no `breadth/` | 2 — breadth research |
| `breadth/` populated, no `L1.md` | 3 — compress each area |
| `L1.md`, no `L2.md` | 4 — cross-compress |
| `L2.md`, no `final.md` | 5 — final output |
| `final.md` | present it |

## Phase 1 — decompose → `areas.md`

Read the codebase or topic. Identify 5 major areas of roughly equal weight and
decide per area whether it needs web research, codebase reading or both —
NEVER force one source on every area.

```markdown
# Areas
1. [Name] - [description] - [source: code|web|both]
```

## Phase 2 — breadth → `breadth/01-name.md` …

Per area, gather ~500 words of raw findings: read files and grep patterns for
`code`; WebSearch/WebFetch for context, state of the art and comparisons for
`web`; both for `both`. No compression yet. Run areas sequentially and save
after each.

## Phase 3 — compress each area → `L1.md`

Per area, three levels, each shorter than the last:

```markdown
## [Area Name]

### Detailed (~200 words)
[What it does, why it exists, key decisions]

### Summary (~50 words)
[One paragraph]

### Essence (3-5 words)
**[Core of this area]**
```

## Phase 4 — cross-compress → `L2.md`

From the five essences and summaries:

```markdown
# Cross-Area Analysis

## Patterns (~100 words)
[Trade-offs, architectural decisions, recurring themes]

## Summary (~30 words)
[One paragraph for the whole]

## Essence (3-5 words)
**[What this IS]**
```

## Phase 5 — final → `final.md`

```markdown
# Distillation: [Subject]

## Essence
**[3-5 words]**

## Summary
[One paragraph from L2]

## Patterns
[Cross-cutting from L2]

## By Area

### [Area 1]: [3-5 word essence]
[Summary from L1]
```

Present it. Done.

## Rules

- ALWAYS read the existing state before starting and save to `.distill/`
  after each phase.
- NEVER compress without reading the source first.
- NEVER repeat a phrase between compression levels; ALWAYS make each level
  shorter than the previous.
- ALWAYS make the essence quotable and true. No marketing language.

From kronael/tools, Unlicense.

More from kronael/tools

Signals

Stars
6
Forks
3
Open issues
0
Last commit (this folder)
Oct 22, 2026 (9 days ago)

From GitHub (fetched Oct 24, 2026).

Details

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

More from kronael

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 distill 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 distill?

CallGet help
Cite this page

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