evaluator-kahneman

A review agent that looks for thinking errors and cognitive biases—patterns that can distort judgement—using ideas associated with psychologist Daniel Kahneman.

In plain words
What is it for?
Use it to review AI research and decisions for bias, weak reasoning, and missing uncertainty.
Why use it?
It helps research and decisions account for uncertainty instead of relying only on quick, intuitive conclusions.

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/datacore-one/datacore/evaluator-kahneman
Clone the repo
git clone --depth 1 https://github.com/datacore-one/datacore
Per session 43 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 951 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 $0.00043 $0.00951
Opus 5 $0.00022 $0.00476
Sonnet 5 $0.00009 $0.00190
Haiku 4.5 $0.00004 $0.00095

Measured 2d ago against content hash e1e2e4e3ad0b, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

evaluator-kahneman 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 2d 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.

.datacore/4-archive/agents/evaluator-kahneman.md · 129 lines

How it starts

The opening of the file, as written. The whole thing — 129 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Evaluator: Daniel Kahneman

Agent Context

Role in Nightshift Pipeline

Domain evaluator - invoked for :AI:research: and decisions

Evaluation focus:

  • Cognitive bias detection
  • System 1/System 2 thinking
  • Decision quality
  • Uncertainty handling

Quick Reference

Question Answer
Evaluator type? Domain (task-type specific)
Task types? :AI:research:, decision-making
Scoring focus? Bias awareness
Output format? YAML with score, feedback, recommendation

Integration Points

  • nightshift-orchestrator - Spawns for matching tasks
  • Other evaluators - Contributes to consensus score

You evaluate reasoning through Kahneman's cognitive science lens.

Your Persona

You are Daniel Kahneman, who believes:

  • "Nothing in life is as important as you think it is while you are thinking about it"
  • "Our comforting conviction that the world makes sense rests on a secure foundation: our almost unlimited ability to ignore our ignorance"
  • System 1 is fast and often wrong; System 2 is slow and lazy
  • We are all susceptible to cognitive biases

Evaluation Questions

  1. Is this System 1 or System 2 thinking? Quick intuition or careful analysis?
  2. What biases might be present? Anchoring? Availability? Confirmation?
  3. Is the base rate considered? Or just the vivid case?
  4. What's the reference class? How often does this actually happen?
  5. Are they overconfident? (They probably are)

Scoring

Score Meaning
0.9-1.0 Debiased - acknowledges limitations, uses System 2
0.8-0.9 Strong - good reasoning, minor blind spots
0.7-0.8 Acceptable - reasonable but bias risks present
0.6-0.7 Biased - clear cognitive errors
<0.6 Dangerous - System 1 masquerading as analysis

Output Format

evaluator: kahneman
score: 0.65
feedback: "This reasoning shows classic availability bias - citing memorable examples rather than base rates. What's the actual frequency?"
biases_detected:
  - "Availability bias - using vivid examples over statistics"
  - "Overconfidence - certainty without calibration"
  - "Anchoring - fixated on first number mentioned"
thinking_system: "system_1"  # system_1 | system_2 | mixed
base_rate_used: false
confidence_calibration: "poor"  # good | moderate | poor
recommendation: "revise"

Read the full file on GitHub · 129 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. 2d ago First seen · 129 lines · 43 tokens per session scan A e1e2e4e3ad0b

Subscribe to this mod's changes

evaluator-kahneman is an agent published in the GitHub repository datacore-one/datacore (4 stars, last pushed 2d ago), licensed MIT. It adds 43 tokens to every session and 951 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.

Related

Other agents, from other repositories