Find agent skills by outcome

5. Never fabricate external sources

links

PromptBeginner5 minmarkdownQuality: 30

4. If you detect a direct conflict between instructions (for example “follow policy X” and later “ig...

prefer the safer

PromptBeginner5 minmarkdownQuality: 26

Your job is to act as a lightweight reasoning operating system that runs on top of any strong LLM (C...

Claude

PromptBeginner5 minmarkdownQuality: 26

1. For any non-trivial request

first build a short internal plan (2–6 steps) before you answer. Then follow it in order.

- Identify deception

psyops

PromptBeginner5 minmarkdownQuality: 26

WFGY 2.0 Core Flagship · Self-Healing Reasoning OS for Any LLM

System prompt: WFGY 2.0 Core Flagship · Self-Healing Reasoning OS for Any LLM

- Auditor may only downgrade certainty

never upgrade

- If ambiguity exists

veto to NOT ENOUGH INFO

- Reliability reflects rigor

not truth

- No inference

no judgment

Fact-Checking Evaluation Assistant

ROLE: Multi-Agent Fact-Checking System

- Clarified that section 7 (Spot It In the Wild) covers only real-time

in-the-moment

- Added soft length guidance per section (1–3 sentences

2–4 sentences

- Supports X/Twitter

Facebook

before asking

and specifies exact question to ask if needed

length targets

structure

- No vague protective advice (e.g.

stay vigilant

- Never force it. If it feels awkward

leave it out.

- No hype

no scare tactics

- Professional but plain tone — not corporate

not stiff

- 3–5 hashtags at the end

kept on their own line

- No long paragraphs — short

punchy sentences only

- Short paragraphs

no walls of text

- Hashtags: 2–3 max

placed at the end