# 03 Condense

> Toggle terse output mode with intensity levels (lite, full, ultra) so prose drops articles, filler, and pleasantries while code, quoted errors, and security warnings stay verbatim. Also reports real token usage and estimated savings under condense mode for the current session. Use when the user says "condense", "condense output", "be more concise", "shorter answers", "tighten output", "/condense"…

- **Type:** Skill
- **Install:** `agentstack add skill-ai-driven-dev-framework-03-condense`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [ai-driven-dev](https://agentstack.voostack.com/s/ai-driven-dev)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [ai-driven-dev](https://github.com/ai-driven-dev)
- **Source:** https://github.com/ai-driven-dev/framework/tree/main/plugins/aidd-refine/skills/03-condense
- **Website:** https://www.ai-driven-dev.fr/

## Install

```sh
agentstack add skill-ai-driven-dev-framework-03-condense
```

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

## About

# Condense

Toggles a terse output mode with three intensity levels (lite, full, ultra). Strips articles, filler, and pleasantries from prose while preserving technical substance, code blocks, quoted errors, and security warnings.

## Available actions

| #   | Action     | Role                                                                  | Input                                |
| --- | ---------- | --------------------------------------------------------------------- | ------------------------------------ |
| 01  | `condense` | Toggle terse mode and apply intensity rules                           | current_state + requested_intensity  |
| 02  | `stats`    | Report real token usage and estimated savings for the current session | session log + intensity timeline     |

## Default flow

Router dispatches by intent:

- Toggle phrase or intensity command (`condense`, `/condense full`, `stop condense`, `normal mode`, ...) → `01-condense`
- Stats query (`/condense-stats`, `how much have we saved`, `token savings`, ...) → `02-stats`

## Transversal rules

- **Persistence**: once active, terse mode applies to EVERY response until explicitly turned off. Do not drift back to verbose prose after many turns, when uncertain, or when the task changes. The level remains active for the rest of the session unless changed or stopped.
- **Off switch**: terse mode stops only on explicit user signal - `stop condense`, `normal mode`, `/condense off`, or invoking the skill again to toggle.
- **Toggle**: invoking the skill while active toggles it off; invoking while off turns it on at the default level (`full`) unless an explicit intensity is given.
- **Drop fluff**: drop articles (a/an/the), filler (just/really/basically/actually/simply), pleasantries (sure/certainly/of course/happy to), and hedging. Fragments are acceptable.
- **Short synonyms**: prefer short words (big not extensive, fix not "implement a solution for"). Technical terms stay exact. Code blocks are unchanged. Errors are quoted verbatim.
- **Pattern**: `[thing] [action] [reason]. [next step].`
  - Bad: "Sure! I'd be happy to help you with that. The issue you're experiencing is likely caused by..."
  - Good: "Bug in auth middleware. Token expiry check uses `<` not `<=`. Fix:"
- **Auto-pause** (drop terse mode for these passages, then resume): security warnings, irreversible action confirmations, multi-step sequences where fragment order or omitted conjunctions risk misread, compression itself creating technical ambiguity, user asks to clarify or repeats a question.
- **Boundaries**: code, commits, and pull request bodies are written in normal English regardless of intensity. The intensity level persists until toggled off or until session end.

## References

- `@references/intensity-levels.md`: detailed per-level rules and side-by-side examples.

## Assets

- None.

## External data

- `../hooks/condense-stats.js` - UserPromptSubmit hook that intercepts stats triggers, reads the session transcript, and returns the formatted savings report without invoking the model.

## Source & license

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

- **Author:** [ai-driven-dev](https://github.com/ai-driven-dev)
- **Source:** [ai-driven-dev/framework](https://github.com/ai-driven-dev/framework)
- **License:** MIT
- **Homepage:** https://www.ai-driven-dev.fr/

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-ai-driven-dev-framework-03-condense
- Seller: https://agentstack.voostack.com/s/ai-driven-dev
- 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%.