feat: stabilized org-agent two-way communication and UX

- Fixed kernel-to-Emacs communication bridge.
- Resolved boot-time crashes in multiple skeletal skills.
- Refined Chat skill prompt to eliminate conversational filler.
- Updated Emacs UI to automatically clean up status markers.
- Synchronized all fixes via Literate Org-mode documents.
- Verified physical two-way interaction via simulation.
This commit is contained in:
2026-04-03 17:25:01 -04:00
parent 93f9ccee17
commit fdb55c616d
30 changed files with 654 additions and 100 deletions

View File

@@ -4,56 +4,53 @@
#+FILETAGS: :llm:provider:gemini:google:psf:
* Overview
The *Gemini Provider Agent* integrates Google's Gemini API as a pluggable System 1 (neural) backend. This skill enables modular updates to the Google backend while maintaining strict architectural alignment with the PSF.
* Phase A: Demand (PRD)
:PROPERTIES:
:STATUS: FROZEN
:END:
** 1. Purpose
Define the interface for reliable communication with the Google Gemini v1beta API.
** 2. User Needs
- *API Integration:* Implementation of the Gemini `contents.parts` protocol.
- *Reliability:* Secure retrieval of endpoints and API keys from the environment.
- *Modularity:* Registration under `:gemini-official` for multi-provider support.
** 3. Success Criteria
*** TODO API Authentication via URL Key
*** TODO Payload Construction
*** TODO Response Parsing Verification
The *Gemini Provider Agent* integrates Google's Gemini API as a pluggable System 1 (neural) backend.
* Phase B: Blueprint (PROTOCOL)
:PROPERTIES:
:STATUS: SIGNED
:END:
** 1. Architectural Intent
Interfaces for executing neural completion requests via Google's generative AI endpoints.
** 2. Semantic Interfaces
#+begin_src lisp
(defun execute-gemini-v1-request (prompt system-prompt)
"Executes a completion request via the Gemini API.")
(defun execute-gemini-request (prompt system-prompt)
"Executes a completion request via the Google Gemini API.")
#+end_src
* Phase D: Build (Implementation)
** Request Execution
#+begin_src lisp :tangle projects/org-skill-provider-gemini/src/provider-logic.lisp
(defun execute-gemini-v1-request (prompt system-prompt)
(let ((api-key (uiop:getenv "LLM_API_KEY"))
(endpoint (uiop:getenv "LLM_ENDPOINT")))
(unless api-key (return-from execute-gemini-v1-request "ERROR: Key missing"))
;; Physical API call logic (mocked for refactor)
(format nil "Executing Gemini request via ~a" endpoint)))
#+begin_src lisp
(defun execute-gemini-request (prompt system-prompt)
(let* ((auth (org-agent:get-provider-auth :gemini))
(api-key (getf auth :api-key))
(bearer-token (getf auth :bearer-token))
(endpoint (or (getf auth :endpoint)
"https://generativelanguage.googleapis.com/v1beta/models/gemini-pro:generateContent")))
(unless (or api-key bearer-token)
(return-from execute-gemini-request "(:type :LOG :payload (:text \"Authentication missing for Gemini\"))"))
(let* ((url (if api-key (format nil "~a?key=~a" endpoint api-key) endpoint))
(headers `(("Content-Type" . "application/json")
,@(when bearer-token `(("Authorization" . ,(format nil "Bearer ~a" bearer-token))))))
(body (cl-json:encode-json-to-string
`((contents . ((parts . ((text . ,(format nil "~a~%~%Prompt: ~a" system-prompt prompt))))))))))
(handler-case
(let* ((response (dex:post url :headers headers :content body))
(json (cl-json:decode-json-from-string response)))
(cdr (assoc :text (cdr (assoc :parts (car (cdr (assoc :parts (car (cdr (assoc :candidates json)))))))))))
(error (c)
(format nil "(:type :LOG :payload (:text \"Neural Engine Failure: ~a\"))" c))))))
#+end_src
* Registration
#+begin_src lisp
;; Register with the kernel
(org-agent:register-neuro-backend :gemini #'execute-gemini-request)
(defskill :skill-provider-gemini
:priority 100
:priority 90
:trigger (lambda (context) nil)
:neuro (lambda (context) nil)
:symbolic (lambda (action context) action))