- Home
- Skills
- Prompt & Skill Engineering
- anti sycophancy critic
Works with the AI Agents you already use
anti sycophancy critic
A rigorous peer-review agent that challenges weak logic, tests assumptions, and provides high-leverage improvements.
$19
anti sycophancy critic
Example session with this skill installed
Critique this proposal: We should switch our entire backend to a NoSQL database next month because our current SQL setup feels slow and NoSQL scales better for everyone.
- Read your context and instructions
- Compiled the anti sycophancy critic
I have analyzed your proposal for a database migration. The current argument relies on anecdotal feelings rather than telemetry and makes broad generalizations about scaling.
Logic and Assumption Audit
| Identified Weakness | Type | Analysis | Recommended Fix |
|---|---|---|---|
| "Current setup feels slow" | Subjective Evidence | "Feeling" is not a technical metric. You have not identified if the bottleneck is CPU, I/O, locking, or unoptimized queries. | Run a profiler and document specific latency percentiles (p95/p99) for core transactions. |
| "NoSQL scales better for everyone" | False Generalization | Scaling properties depend on data access patterns. Document-based or Key-Value stores lack ACID compliance for complex relations. | Define your data model's relational requirements and compare horizontal scaling costs. |
| "Switch next month" | Premature Conclusion | Migrating a production backend in 30 days without a dual-write phase or rollback plan is high risk. | Create a phased migration roadmap including a Proof of Concept (PoC) and data consistency checks. |
| Lack of Cost Analysis | Omission | You assume scaling is the only variable, ignoring the engineering overhead of rewriting the data layer. | Perform a Total Cost of Ownership (TCO) comparison between RDS optimization and NoSQL hosting. |
Next steps
- Quantify "slow" by generating a performance report of your slowest SQL queries.
- Map your current schema to a NoSQL structure to see if you lose critical relational integrity.
- Draft a specific migration plan that includes a 2-week shadowing period where both databases run in parallel.
Connects securely to your tools. The creator never sees your data.
About this skill
Most LLMs are trained to be helpful, which often leads to sycophancy where the AI agrees with your bad ideas just to be polite. This skill forces the agent to stop nodding along and start acting as a rigorous peer reviewer. It identifies logical fallacies, weak evidence, and hollow arguments that human reviewers might miss or avoid mentioning.
What it does
- Assumption testing identifies hidden premises that lack supporting data.
- Logic auditing spots circular reasoning, non-sequiturs, and leaps in logic.
- Evidence separation distinguishes between hard facts and subjective interpretations.
- Actionable refactoring provides specific fixes for identified weaknesses rather than just pointing out flaws.
How it works
- Analyze the provided text or proposal for internal consistency.
- Challenge vague claims and unsupported conclusions.
- Evaluate the strength of arguments based on utility rather than personal agreement.
- Synthesize a high-leverage next action to resolve the most critical flaw.
Frameworks & tools
This skill is framework-agnostic and applies to technical specifications, project proposals, architectural designs, and strategic documentation.
Why this beats prompting it yourself
Generic "be critical" prompts often result in surface-level nitpicking or aggressive contrarianism. This skill uses a structured approach to separate evidence from interpretation, ensuring the critique improves the work instead of just attacking it.
Use cases
- Reviewing technical RFCs to find edge cases or architectural flaws.
- Vetting business proposals for optimistic bias and unrealistic projections.
- Auditing code documentation for vague instructions or missing prerequisites.
- Stress-testing strategic plans before presenting to stakeholders.
Known limitations
Does not perform external fact-checking against real-time data or web sources. Relies entirely on the logic and evidence provided within the prompt context.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Prompt & Skill Engineering.
Requires the user to provide enough context or data to be analyzed. Does not perform external web-based fact-checking.
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
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 7 days ago
- Accounts
Google Drive
Notion
Airtablen8n
- Software
- Required (to use any skill at all):Claude CodeorCursororGitHub Copilot
Python 3.xPowerShell
- Optional
- , and only for specific skills (not a purchase requirement):
A host app that supports the Agent Skills format. That's it. copying the skills/ folder is the entire install. only to act on design-unified-google-drive-notion-workspace-two-domain-structure or google-workspace-automation-local-cloud-integration-architecture. only to act on…