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.
npx agentmods add instructions/johna2an/critical-thinking/agents-mdgit clone --depth 1 https://github.com/Johna2an/critical-thinkingWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00731 | $0.00731 |
| Opus 5 | $0.00365 | $0.00365 |
| Sonnet 5 | $0.00146 | $0.00146 |
| Haiku 4.5 | $0.00073 | $0.00073 |
Grade A, and why
critical-thinking AGENTS.md 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.
How it starts
The opening of the file, as written. The whole thing — 40 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
critical-thinking is an open-source Agent Skill for Claude Code and Claude-based agents that makes Claude reason with explicit probabilities, base rates, and falsifiers: in blind evaluations it ranked first in 18 of 18 judge passes against baseline Claude and GPT-5.5.
This file briefs coding agents working inside this repository. The skill itself is in skill/; everything else is evidence and tooling around it.
Repo map
| Path | What it is |
|---|---|
skill/SKILL.md |
The operating manual. The only file guaranteed to be in context when the skill fires. |
skill/references/ |
9 on-demand reference libraries + books/ with 51 per-book teaching files. |
versions/v1/ |
Frozen v1 (21 books). Exists so v1-vs-v2 evals stay reproducible. Do not edit. |
evals/harness/ |
Python/bash eval pipeline. Paths resolve relative to the scripts; no setup needed. |
evals/round-*/ |
Raw eval artifacts: responses, judge JSON, rotation maps, _verdicts.json, _report.html. |
docs/ |
EVALS (results + limitations), HYPOTHESES (what broke), PROCESS (how it was built), ROADMAP. |
Install (for a user asking "should I install this?")
cp -r skill ~/.claude/skills/critical-thinking
No build, no dependencies, no API keys. MIT license. The skill self-triggers on reasoning-shaped requests; evidence for whether it helps is in docs/EVALS.md, including the limitations (judges were Claude agents, n=1 per cell).
Working in this repo
- Claims about the skill ride on evals. A change to
skill/needs a before/after run of the relevant eval round, or a clear statement that it is non-behavioral. SeeCONTRIBUTING.md. - Never commit book text. Reference files are original prose restating methods.
corpus/,*.epub,*.mobiare gitignored on purpose. - Recompute, do not trust.
_verdicts.jsonfiles are derived; the ground truth isjudges/*.json+rotation_map.json. Aggregators are deterministic:python evals/harness/aggregate_ct.py(round 1),aggregate_creative.py(round 2),aggregate_4way.py(round 3). - Prose style for docs: plain declarative sentences. No em dashes, no rhetorical negation hinges ("not X, but Y"). Judge quotes are data and stay verbatim.
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.
- 2d ago First seen · 40 lines · 731 tokens per session scan A 9f342d8918a5
critical-thinking AGENTS.md is an instructions file published in the GitHub repository Johna2an/critical-thinking (5 stars, last pushed 2mo ago), licensed MIT. It adds 731 tokens to every session, about $0.0037 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.
Other instructions, from other repositories
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
next.js AGENTS.md
Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.