#+TITLE: SKILL: The Scientist Agent (Universal Literate Note)
#+ID: skill-scientist
#+STARTUP: content
#+FILETAGS: :debugging:science:logic:tdd:psf:
#+DEPENDS_ON: skill-tdd-runner skill-scribe-rca

* Overview
The **Scientist Agent** provides a formal, hypothesis-driven approach to debugging. Instead of "trial and error," it formulates symbolic theories about why a failure occurred, designs experiments (test cases), and updates the system's **Institutional Memory** upon discovery.

* Phase A: Demand (PRD)
:PROPERTIES:
:STATUS: FROZEN
:END:

** 1. Purpose
Eliminate speculative debugging through rigorous scientific methodology.

** 2. User Needs
- **Hypothesis Formulation:** Neural generation of potential failure causes.
- **Experimental Design:** Autonomous creation of minimal failing test cases.
- **Theory Verification:** Execution of tests via the TDD Runner.
- **Knowledge Update:** Permanent update to `RCA.org` to prevent regression.

* Phase D: Build (Implementation)

** Scientific Loop
#+begin_src lisp :tangle projects/org-skill-scientist/src/scientist-logic.lisp
(defun scientist-formulate-hypothesis (failure-log)
  "Analyzes an error and proposes a 'Theory of Failure'."
  (org-agent:ask-neuro 
   (format nil "Explain why this failure occurred and propose a specific experiment to prove it: ~a" failure-log)
   :system-prompt "You are a PSF Senior Debugging Scientist. Use formal logic and the scientific method."))

(defun scientist-run-experiment (hypothesis)
  "Designs a minimal test case based on a hypothesis."
  ;; Delegates to TDD Runner to create and run the experiment.
  (org-agent:spawn-task (format nil "Create a minimal failing test for: ~a" hypothesis)))
#+end_src

* Registration
#+begin_src lisp
(defskill :skill-scientist
  :priority 90
  :trigger (lambda (context) (search "ERROR" (getf (getf context :payload) :text)))
  :neuro #'scientist-formulate-hypothesis
  :symbolic (lambda (action context) action))
#+end_src