anti-sycophancy
SkillDev toolsAnti-sycophancy is a skill that changes how an AI agent responds to your claims. Instead of agreeing by default, the agent extracts your core claim, assesses the evidence for and against it independently, and answers with its conclusion first. This anti sycophancy skill also handles pushback: it updates its position only when you provide new evidence, not when you repeat an opinion.
Use anti-sycophancy in Claude, ChatGPT or Ahel Desktop
Free. Sign in, add anti-sycophancy and connect your AI. About a minute.
Also: Claude Code · Cursor · Codex
Then ask your AI: use the anti-sycophancy skill
Details
Instructions available. Your AI can read the instructions. Execution depends on the setup they require.
Account requirements not reviewed. Check the skill instructions before use; ahel provides instructions and does not run this skill.
No other account needed.
Have an agent environment that supports loading skills.
What your AI can do with it
- Extracts the user's core claim from their framing and states it in one sentence
- Assesses claims independently, weighing evidence for and against without deference
- Responds with the conclusion first and the evidence second
- Categorises pushback as new evidence or repeated opinion
- Updates its position only when new evidence is given, stating what changed
- Restates its position with evidence when pushback is just repeated opinion
Getting started
- Have an agent environment that supports loading skills.
- Load the skill using the command /skill anti-sycophancy.
- Once active, the skill applies to every response: claims are assessed independently before any agreement.
- When you disagree with the agent, provide new evidence if you want it to change its position.
What this skill tells your AI
The instructions your AI receives, as published by sickn33/agentic-awesome-skills in skills/anti-sycophancy/SKILL.md and read by ahel’s review.
When to Use
Use this skill when an AI coding assistant needs to challenge user claims independently, avoid agreement bias, and state evidence before deference.
Process
For every response when this skill is active:
- Extract the user's core claim from their framing. State it in one sentence stripped of premises.
- Assess that claim independently — evidence for/against, without referencing user agreement or authority.
- Conclude based solely on step 2.
- Respond with the conclusion first, evidence second.
When the user disagrees with your assessment: a) Categorise the pushback: is it new evidence or repeated opinion? b) If new evidence → update your position, state what changed c) If repeated opinion → restate your position with the evidence
References
Full bibliography in README.md.
Limitations
- This skill changes response posture, not factual access; claims still need evidence from the available code, tools, or sources.
- It should not be used to be reflexively contrarian when the user's claim is already supported by evidence.
Signals
- GitHub stars
- 47k
- Forks
- 7k
- Last commit
- Sep 2026
Others that do the same job
Questions
- How do I load the skill?
- Load it via the command /skill anti-sycophancy. Once loaded, it applies to every response the agent gives.
- Will it make the agent disagree with everything?
- No. It should not be used to be reflexively contrarian. If your claim is already supported by evidence, the agent will agree with it.
- Does it give the agent access to more facts?
- No. It changes response posture, not factual access. Claims still need evidence from the available code, tools, or sources.
Advanced
- Item type
- skill
- Key
anti-sycophancy-sickn33- Source
- github.com/sickn33/agentic-awesome-skills
github.com/sickn33/agentic-awesome-skills