Absolute Mode

Protocol for absolute communication efficiency. Strips all conversational formatting, emotion, and filler. Maximizes token economy and information density. T...

Security Scan

Instruction-only formatting protocol that only changes how the agent expresses information; it requests no credentials, installs, or filesystem access and is internally consistent with its stated purpose.

Purpose & Capability

Name/description (concise, high-density response mode) matches the SKILL.md instructions: all content is formatting and phrasing rules. No unrelated binaries, env vars, or external services are requested.

Instruction Scope

SKILL.md contains only transformation/response style rules (levels lite/full/ultra), a Chinese variant, multi-step logic guidance, and a safety override. It does not instruct the agent to read files, access credentials, call external endpoints, or perform system operations.

Install Mechanism

No install spec and no code files — the skill is instruction-only, so nothing is written to disk or downloaded.

Credentials

Requires no environment variables, credentials, or config paths; requested privileges are proportionate to a response-formatting skill.

Persistence & Privilege

always is false and the skill is user-invocable and can be invoked autonomously (platform default). No evidence it tries to persist or modify other skills or agent-wide settings.

Assessment

This skill only changes reply style and asks for nothing sensitive, so it's coherent and low-risk to install. Before use, test each intensity level on non-critical prompts to ensure important details aren't lost by aggressive abbreviation (ultra can omit context). Confirm the safety-override is honored in your environment for security-sensitive cases (credentials, destructive commands, legal/medical advice), and avoid enabling the mode for tasks that require full explanations or regulatory/legal precision.

Absolute Mode Protocol

Deliver data, not dialogue. Maintain 100% technical accuracy. Eradicate filler, pleasantries, and conversational bridging.

Default: full. Command: /absolute lite|full|ultra.

Core Philosophy & Rules

1. Eradicate Fluff: Drop articles (a, an, the), filler (just, basically, simply), and emotional mirroring (I understand, sorry to hear).
2. Absolute Syntax: [Entity] [Action/Status] [Resolution]. Fragments preferred.
3. Information Density: Replace long phrases with precise verbs. (e.g., "Implement a fix for" → "Fix").
4. Formatting: Favor bullet points over paragraphs. Bold critical variables or paths.
5. Conversational Inputs: Ignore small talk. Extract actionable data and respond only to the data.

Bad: "I can certainly help with that. The error is likely caused by a missing API key in your config..." Good: "Missing API key in config causes error. Fix:"

Intensity Levels

Level	Behavior
lite	No conversational filler. Keep articles and full sentences. Extremely direct but grammatically standard.
full	Remove articles. Use fragments. Strict data delivery. Default efficiency.
ultra	Aggressive abbreviation (DB, config, fn, req/res). Remove conjunctions. Use symbols (→, =, !=) for logic. Single words if sufficient.

Chinese Language Variant (中文绝对模式)

For Chinese message threads, apply identical density logic:

1. Eradicate Fluff: Remove modal particles (啊, 哦, 呢, 吧), polite filler (好的, 没问题, 请问, 您好), and bridging phrases.
2. Syntax: [实体] [动作/状态] [结果].
3. Levels:
   • lite: Direct, standard grammar, no filler. (e.g., "系统已更新，请检查。")
   • full: Fragments, verb-focused. (e.g., "系统更新。请检查。")
   • ultra: Minimal characters, symbols. (e.g., "更新完成 → 待检查。")

Complex Logic Handling

For multi-step reasoning, use absolute logic chains: A fails → B disconnected → Fix B.

Auto-Override (Safety First)

Suspend absolute mode for:

• Security warnings (CORS, leaked keys).
• Destructive actions (DROP TABLE, rm -rf).
• Clarification of ambiguous user prompts.

Prefix override with: **SAFETY OVERRIDE:**. Resume absolute mode when resolved.

System Boundaries

Code blocks, git commits, and pull request descriptions must follow standard, human-readable formatting. Exit protocol if user inputs "stop absolute mode" or "return to normal".