@dzhechkov/harness-cli 0.3.73 → 0.3.74

package/README.md

@@ -408,30 +408,39 @@ dz dashboard    # Visual panel with all packages, adapters, skill packs
 ## Example: Thesis Defense Preparation (Academic Preset)
 
 ```bash
-# Install academic skills (5 skills for thesis defense committee)
-dz init --target claude-code --preset academic
+# Install with AgentDB (remembers patterns across students):
+dz setup --target claude-code --preset academic --memory agentdb
 
-# Open Claude Code in the folder with student materials and use:
-#   "Analyze this thesis against project format criteria"
-#   "Generate 6 defense questions of varying difficulty"
-#   "Check document package completeness"
+# Or lightweight:
+# dz init --target claude-code --preset academic
 ```
 
-Five skills auto-activate in Claude Code:
+**Prepare:** Create a folder per student with thesis.pdf + review.pdf + external-review.pdf + antiplagiat.pdf.
 
-| Skill | What it does |
-|-------|-------------|
-| `document-checker` | Validates package completeness: thesis, supervisor review, external review, presentation |
-| `dissertation-review` | Evaluates thesis against ГЭК criteria (research/project format), generates summary |
-| `question-generator` | Generates 6 questions (basic → critical) with expected answer keywords |
-| `defense-evaluator` | Analyzes live defense transcript — structure, topic coverage, delivery quality |
-| `answer-assessor` | Evaluates Q&A answers — completeness, depth, alignment with reviewer remarks |
+**Pre-defense** (open Claude Code in student folder):
+```
+"Check document package completeness"     → document-checker
+"Analyze this thesis"                     → dissertation-review (format, criteria, grade)
+"Generate 6 defense questions"            → question-generator (basic → critical, page refs)
+```
 
-**Live defense support:** Feed real-time transcript from speech-to-text (e.g., Whisper + VB-Cable on Windows) → defense-evaluator and answer-assessor analyze as the defense progresses.
+**During defense** (feed live transcript via [Whisper](https://github.com/openai/whisper) + [VB-Cable](https://vb-audio.com/Cable/)):
+```
+"Analyze this defense transcript"         → defense-evaluator (structure, coverage, delivery)
+"Evaluate the student's answers"          → answer-assessor (completeness, depth, reviewer alignment)
+```
 
-Skills contain **only evaluation criteria and methodology** — no student data.
+| When | Skill | What it does |
+|------|-------|-------------|
+| Before | `document-checker` | Package completeness: thesis, reviews, antiplagiat |
+| Before | `dissertation-review` | ГЭК criteria, research/project format, grade 1-10, team project check |
+| Before | `question-generator` | 4-6 questions with page refs and expected keywords |
+| During | `defense-evaluator` | Live transcript → structure, coverage, delivery quality |
+| During | `answer-assessor` | Q&A evaluation → completeness, depth, reviewer remarks |
+
+**Key features:** TO BE vs data detection, LTV/CAC > 10 warning, reviewer score divergence alerts, team project evaluation. With AgentDB, patterns persist across students.
 
-Recommended workflow: `document-checker` → `dissertation-review` → `question-generator` → `defense-evaluator` → `answer-assessor`
+Skills contain **only evaluation criteria and methodology** — no student data.
 
 ---
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@dzhechkov/harness-cli",
-  "version": "0.3.73",
+  "version": "0.3.74",
   "description": "The dz CLI — install AI skills for Claude Code, Codex, OpenCode, Hermes. 11 commands, 7 presets, 4 platform adapters.",
   "type": "module",
   "license": "MIT",