Skip to content
Skills Library
AI & agents · Automation

compression

/compression

Use when the user asks to reduce token output, compress responses, switch to terse mode, or invokes /siftcoder:compress. Wraps the optional sift-compress companion plugin if installed; otherwise defines the rule directly.

Repository
sift-coder ↗
Licence
MIT
Last repository change
03.06.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 03.06.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Use when the user asks to reduce token output, compress responses, switch to terse mode, or invokes /siftcoder:compress. Wraps the optional sift-compress companion plugin if installed; otherwise defines the rule directly.

Original author description

Use when the user asks to reduce token output, compress responses, switch to terse mode, or invokes /siftcoder:compress. Wraps the optional sift-compress companion plugin if installed; otherwise defines the rule directly.

How it works

  1. Modes
  2. Rules
  3. Mode behavior
  4. Boundaries
  5. Handoff template
  6. Persistence
  7. Companion plugin
Original SKILL.md headings · 84 lines, 379 words
SKILL.md preview
---
name: compression
description: Use when the user asks to reduce token output, compress responses, switch to terse mode, or invokes /siftcoder:compress. Wraps the optional sift-compress companion plugin if installed; otherwise defines the rule directly.
---

# Compression mode

Lean output mode with explicit information preservation. This is not "be vague and short." It compresses wording while preserving decisions, commands, file paths, errors, risks, and next actions.

## Modes

- **lite** — drop pleasantries, hedging
- **full** — drop articles, filler, hedging, pleasantries (default)
- **ultra** — fragments encouraged, headers minimal
- **dense** — maximum signal per line; keep all evidence, scores, and caveats
- **handoff** — compact project state for another agent/session: goal, decisions, files, risks, next steps
- **commit** — apply only when writing commit messages
- **review** — apply only during code review
- **off** — normal style

## Rules

Drop:

- Pleasantries and meta narration
- Hedge words unless uncertainty matters
- Repeated caveats
- Restating the user's request
- Articles when the sentence remains clear

Keep:

- Technical terms exact
- Code blocks unchanged
- Error messages exact
- File paths exact
- Version numbers exact
- Risk/impact/likelihood labels
- Commands and flags
- User-facing decisions and tradeoffs

Fragment OK. Short synonyms OK (`fix` not "implement a solution for"). Never compress away causality: "because X" often matters.

## Mode behavior

| Mode | Target | Shape |
|---|---|---|
| lite | everyday shorter answers | normal grammar, fewer words |
| full | default compact engineering | short paragraphs, tight bullets |
| ultra | low-bandwidth/status | fragments, no softeners |
| dense | audits/reviews/plans | tables or bullets, all evidence preserved |
| handoff | session transfer | structured state block |
| commit | commit messages only | imperative subject + concise body |
| review | code review only | findings first, no summary padding |

## Boundaries

Never compress:

- Security warnings
- Irreversible action confirmations
- Legal/medical/financial safety caveats
- Multi-step sequences where fragment order risks misread
- User-visible final instructions that must be followed exactly

## Handoff template

Use `handoff` mode for compaction or agent transfer:

```
Goal:
State:
Decisions:
Files touched:
Known risks:
Verification:
Next step:
Do not:
```

## Persistence

Mode persists for the session unless changed (`/siftcoder:compress off`) or session ends.

## Companion plugin

If `sift-compress` is installed as a separate plugin, defer to it (it ships hooks that maintain mode state across compactions). Otherwise, this skill enforces the rule inline.

Versions

RepositoryScoreLicence
This page · ialameh/sift-coder100MIT
thedaviddias/front-end-checklist100MIT
kalyvask/winning-writing100MIT
jacob-balslev/skill-graph100Apache-2.0
seandavi/scriptorium95MIT

Database snapshot: 13.11.2026 · package version be55848f8879 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100markdown-compression
    This skill should be used when the user asks to 'compress markdown', 'shrink this file', 'optimize tokens', 'reduce file size', 'compress instructions', 'make this more concise', 'minimize this prompt', 'compress CLAUDE.md', 'compress ARCHITECTURE.md', 'optimize agent…
  • 100compress
    Use when the user says 'tokenstack', 'compression', 'token savings', 'proxy status', or asks about context window usage.
  • 100text-optimize
    Optimizes text, docs and LLM prompts (CLAUDE.md, rules, agent/skill files) for token efficiency in place, snapshot-backed, without losing facts. Use when asked to optimize, compress or cut tokens of named files. Triggers - optimize, reduce tokens, compress, deep compress, сожми, оптимизируй.
  • 100agents-md-compress
    Use when the user asks to shorten, compress, or deduplicate a repository's `AGENTS.md` without losing rules.
  • 100text-optimize
    Optimizes text, docs and LLM prompts (AGENTS.md, rules, agent/skill files) for token efficiency in place, snapshot-backed, without losing facts. Use when asked to optimize, compress or cut tokens of named files. Triggers - optimize, reduce tokens, compress, deep compress, сожми, оптимизируй.
  • 95token-optimizer
    Maximize Claude's output quality while minimizing input token usage. Use this skill whenever a user wants to compress prompts, reduce token consumption, extract maximum output from Claude, write high-density instructions, optimize system prompts, or improve AI communication…

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.