Phronesis is a practical-wisdom project for applying philosophical and personal judgment frameworks to real situations.
The repository is designed to be opened directly in an agent workspace such as Codex or Claude Code. It is not meant to be installed globally. Project-local agent entrypoints load the relevant module only when a user asks for that lens.
git clone git@github.com:OctopusGarage/phronesis.git
cd phronesis
codexThen ask with the local Codex skill:
$camus-absurd I keep postponing my life for a future goal. How would Camus analyze this?
Chinese questions are supported:
$camus-absurd 我一直为了未来目标牺牲当下生活,加缪会怎么分析?
For Claude Code:
cd phronesis
claudeThen ask with the local Claude skill:
/camus-absurd 我最近总觉得工作只是角色表演,加缪的演员这一章怎么看?
camus-absurd: applies Albert Camus's The Myth of Sisyphus and the user's related reading notes to questions about meaning, freedom, suffering, love, creation, religion, rational limits, social roles, and present living.wang-yangming-xinxue: applies Wang Yangming's xinxue and Chuan Xi Lu to questions about moral judgment, self-cultivation, knowing versus acting, conscience, and self-deception.hume-empiricism: applies David Hume's empiricism and the user's Hume notes to questions about evidence, uncertainty, induction, causality, selfhood, religion, convention, and motivation.rousseau-mill-politics: applies Rousseau on legitimacy and Mill on liberty to questions about authority, popular sovereignty, majority power, harm, and free speech.philosophy: routes broad philosophical questions to the right domain or narrower project module, using the user's philosophy overview note as a knowledge-map source.science-logic-philosophy: applies the user's science-and-logic philosophy framework to questions about logic, evidence, scientific method, models, causality, mathematics, AI, cognition, and study planning.logic-reasoning-fallacies: applies the user's practical logic and fallacy framework to audit arguments, claims, hidden premises, evidence strength, causal reasoning, and cognitive biases.shelly-kagan-death-philosophy: applies Shelly Kagan's Yale Death course and the user's course notes to questions about death, identity, survival, physicalism, immortality, fear, life value, and suicide as a philosophical topic.
modules/: canonical thought modules and reference material..agents/skills/: Codex/OpenAI project skills..claude/skills/: Claude Code project skills and slash-command entrypoints.evals/: evaluation cases, rubrics, and reports.prompts/: optional prompt examples and templates.mcp/: future MCP servers and shared schemas.references/: project-level research and source indexes.scripts/: deterministic maintenance and validation helpers.
This project separates development instructions from thought-module content:
AGENTS.mdandCLAUDE.mddescribe how agents should work inside this repository..agents/skills/and.claude/skills/are thin local entrypoints for agent tools.modules/contains the canonical knowledge and response guidance.evals/checks whether the module answers real scenarios with the intended boundaries and tone.
Use the project by opening the repository directory in an agent tool. Do not copy the module into global agent settings unless you intentionally want it available outside this repository.
scripts/validate.shTo run the Camus module eval suite:
scripts/run-camus-evals.shTo run the Wang Yangming module eval suite:
scripts/run-wang-yangming-xinxue-evals.shTo run the Hume and Rousseau/Mill module eval suites:
scripts/run-hume-empiricism-evals.sh
scripts/run-rousseau-mill-politics-evals.shTo run the general philosophy router eval suite:
scripts/run-philosophy-evals.shTo run the science/logic philosophy eval suite:
scripts/run-science-logic-philosophy-evals.shTo run the logic reasoning and fallacy eval suite:
scripts/run-logic-reasoning-fallacies-evals.shTo run the Shelly Kagan death philosophy eval suite:
scripts/run-shelly-kagan-death-philosophy-evals.sh