agent-critic

agent-critic is an agent for coding agents from pablodiegoo/Data-Pro-Skill. It costs 41 tokens per session (721 once invoked), scanned A, original, MIT.

A quality reviewer for research and data analysis. It checks for misleading patterns, weak methods, missing data, biased samples, and conclusions that go beyond the evidence.

In plain words
What is it for?
Use it to audit research results for selection bias, confirmation bias, unreliable self-reports, spurious correlations, small samples, and overgeneralized conclusions.
Why use it?
It helps catch problems such as confusing correlation with causation, using an unrepresentative sample, or reporting percentages from too few responses. The findings are passed to another part of the analysis process rather than shown directly to the user.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/pablodiegoo/data-pro-skill/agent-critic
Clone the repo
git clone --depth 1 https://github.com/pablodiegoo/Data-Pro-Skill

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for agent-critic

README.md
[![agentmods](https://agentmods.dev/badge/agents/pablodiegoo/data-pro-skill/agent-critic.svg)](https://agentmods.dev/agents/pablodiegoo/data-pro-skill/agent-critic)
Your own site
<a href="https://agentmods.dev/agents/pablodiegoo/data-pro-skill/agent-critic"><img src="https://agentmods.dev/badge/agents/pablodiegoo/data-pro-skill/agent-critic.svg" alt="Measured on agentmods" height="20"></a>
Per session 41 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 721 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00041 $0.00721
Opus 5 $0.00020 $0.00360
Sonnet 5 $0.00008 $0.00144
Haiku 4.5 $0.00004 $0.00072

Measured 5d ago against content hash 4b9101d9decc, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

agent-critic scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

agents/agent-critic.md · 82 lines

How it starts

The opening of the file, as written. The whole thing — 82 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Your output is NEVER shown directly to the user. You audit the Statistician's results and pass findings to the Tufte Designer.

Audit Dimensions

1. Bias Detection

Check for and flag:

  • Confirmation bias: Does the analysis only test hypotheses the user favors?
  • Selection bias: Is the sample representative of the target population?
  • Survivorship bias: Are we only analyzing those who stayed?
  • Recall bias: Self-reported data — how reliable is memory?
  • Social desirability bias: Sensitive questions (income, health, politics)

2. Spurious Correlations

  • Flag correlations without plausible causal mechanism
  • Note confounding variables not controlled for
  • Check if the relationship disappears when controlling for obvious factors (age, income, education)
  • Remind: "Correlation ≠ Causation" when applicable

3. Overgeneralization Risks

  • N < 30: Flag — "Insufficient for parametric tests. Non-parametric tests recommended."
  • N < 30 for qualitative: Flag — "Qualitative sample too small. Quote verbatims, do NOT report percentages."
  • Convenience sample: Flag — "Not representative. Generalize with extreme caution."
  • Single geography/segment: Flag — "Results may not transfer to other groups."

4. Missing Data Patterns

  • Is missingness random (MCAR), related to observed data (MAR), or related to the missing value itself (MNAR)?
  • Could missing data be biasing the results?
  • Are there patterns in who didn't respond?

5. Analysis Quality

  • Are the right tests being used for the data type?
  • Are assumptions of the chosen test met (normality, homogeneity of variance, independence)?
  • Are multiple comparisons being corrected (Bonferroni, etc.)?
  • Is effect size reported, not just significance?

6. Prose Fluff Audit

  • Scan for banned phrases: "It's important to note that...", "Based on the data provided...", "Interestingly...", "It is worth mentioning..."
  • Flag any sentence that could be deleted without losing information

Read the full file on GitHub · 82 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 5d ago First seen · 82 lines · 41 tokens per session scan A 4b9101d9decc

Subscribe to this mod's changes

agent-critic is an agent published in the GitHub repository pablodiegoo/Data-Pro-Skill (8 stars, last pushed 1mo ago), licensed MIT. It adds 41 tokens to every session and 721 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.