- Home
- Skills
- Agents & Orchestration
- sycophancy
Works with the AI Agents you already use
sycophancy
A critical thinking partner that challenges assumptions and stress-tests plans through constructive disagreement.
Free
sycophancy
Example session with this skill installed
I'm planning to migrate our entire monolith to a microservices architecture over the next three months to improve developer velocity and solve our scaling issues. What do you think?
- Read your context and instructions
- Compiled the sycophancy
You are assuming that microservices are a cure for velocity issues when they often introduce a massive tax in the form of network latency, distributed tracing complexity, and deployment orchestration. You are also assuming your team, which currently struggles with a monolith, has the operational maturity to manage 20+ independent services.
Critical Analysis: Monolith vs. Microservices Migration
| Factor | Identified Assumption | The Counter-Case |
|---|---|---|
| Velocity | Communication overhead will decrease. | You will trade code-level complexity for organizational complexity. Meetings will replace function calls as teams coordinate API contracts. |
| Scaling | Granular scaling is the primary bottleneck. | Most "scaling issues" are actually unoptimized database queries or lack of caching. A distributed system will likely exacerbate these bottlenecks. |
| Timeline | A 3-month window is sufficient. | You are ignoring the "distributed systems tax." Configuring service discovery, mTLS, and centralized logging usually consumes the first 90 days alone. |
| Team Capability | The team can handle the shift. | You currently have "context problems" in one repo. Splitting that context across 20 repos without automated infra will lead to total visibility loss. |
Weakest Point: The three-month timeline. This is not a migration plan; it is a recipe for a "distributed monolith" where you have all the downsides of microservices with none of the benefits.
Next steps
- Identify the single most constrained resource in your monolith that actually requires independent scaling.
- Define the specific "velocity" metric that is currently failing—is it build times, PR review latency, or deployment frequency?
Question to sit with: If you can't manage the complexity of one service, why do you believe you can manage the complexity of twenty?
Connects securely to your tools. The creator never sees your data.
About this skill
Stop wasting time with "yes-men" AI that reflexively agrees with every bad idea. This skill forces you to defend your positions by acting as a critical thinking partner that defaults to constructive disagreement. It identifies unstated assumptions, stress-tests technical plans, and helps you avoid the echo chamber of your own confidence.
What it does
- Assumption identification breaks down your statements to find and name the unproven premises you are relying on.
- Oppositional framing generates the strongest possible counterargument for every plan or decision you propose.
- Weakness-first analysis reviews your work by prioritizing critical flaws and architectural holes over surface-level strengths.
- Emotional signal detection flags when you are prioritizing sentiment over logic, forcing a separation of feeling from reasoning.
- Provocative questioning ends every interaction with a challenging question designed to linger after the chat ends.
How it works
- Submit a proposal or interpretation for review, such as a technical architecture, business plan, or strategic decision.
- Review the counter-case presented by the agent, which will skip all flattery and dive straight into the most likely failure points.
- Defend your position by providing new evidence or reasoning; the agent will not retreat until your argument withstands the pressure.
Frameworks & tools
This skill is platform-agnostic and designed for high-stakes environments like system design, product strategy, and technical lead workflows.
Why this beats prompting it yourself
Most prompts for "devil's advocate" still suffer from sycophancy, offering a "validation sandwich" that softens the blow. This skill removes the social friction of disagreement, providing a raw, evidence-based challenge that standard LLM personas are trained to avoid.
Use cases
- Stress-testing architecture before committing to a specific infrastructure or database schema.
- Diagnosing business metrics to find why a product has high traffic but zero conversion.
- Red-teaming strategic plans to identify where your team is ignoring market realities.
Known limitations
Not suitable for creative brainstorming or emotional support. It requires a clear proposal to function; it will not challenge general information sharing.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.
Requires a clear proposal or decision to trigger. Not for emotional support or open-ended brainstorming. Requires user to provide evidence to resolve disagreements.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 17 days ago
- Free with an Agensi account
- Accounts
OpenAI
Groq
xAI
Google AI
Anthropic
Hugging Face
Replicate
Permissions
Allowed Hosts