# Statement Compression

> >

- **Type:** Skill
- **Install:** `agentstack add skill-npbuilds-skill-library-statement-compression`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [npbuilds](https://agentstack.voostack.com/s/npbuilds)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [npbuilds](https://github.com/npbuilds)
- **Source:** https://github.com/npbuilds/skill-library/tree/main/skills/binding-vow/statement-compression
- **Website:** https://skill-library-prod.web.app/

## Install

```sh
agentstack add skill-npbuilds-skill-library-statement-compression
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Statement Compression — Director

A correct statement in the wrong shape still fails. This director realizes six-eyes' Phase 5
(Compress): it routes a clean, audited statement to the compression format that fits its
audience tag. Compression is force — giving up scope to gain precision — so the audience tag
from `problem-diagnosis` is the controlling input here.

## Routing Table

| Audience / need | Child skill | Output |
|---|---|---|
| exec — fastest possible | `bluf-shaper` | Bottom-Line-Up-Front: one-sentence answer, minimal background, priority-ordered discussion |
| exec / public — substantial doc | `executive-distiller` | Full recursive Minto Pyramid (calls writing/rhetoric/argument-structure) |
| peer / self — narrative intro | `scqa-formatter` | Situation / Complication / Question / Answer |
| LLM — prompt | `cursed-speech` | Anthropic canonical order: role → context → data → examples → numbered instructions → output format → self-check |
| (reference) | `minto-scqa` | Reference: Minto Pyramid, SCQA, MECE, BLUF |

## Routing Logic

1. Read the audience tag from the diagnosis profile. It maps directly to the format above.
2. For LLM audiences, `cursed-speech` additionally recommends downstream skills the formulated
   prompt should invoke — surface those to the user.
3. Pull `minto-scqa` whenever a writer needs the *why* behind the format choice.
4. Hand the compressed statement back to `six-eyes` for the final `statement-grader` pass.

## Scope Boundaries

- **In scope:** shaping an already-audited statement for a target audience.
- **Out of scope:** auditing the statement's logic (→ `statement-audit`), choosing the audience
  (→ `problem-diagnosis`). Compress *after* audit, never before — compressing a flawed statement
  just hides the flaw.

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [npbuilds](https://github.com/npbuilds)
- **Source:** [npbuilds/skill-library](https://github.com/npbuilds/skill-library)
- **License:** MIT
- **Homepage:** https://skill-library-prod.web.app/

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-npbuilds-skill-library-statement-compression
- Seller: https://agentstack.voostack.com/s/npbuilds
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.