Add ECC Harness: phd_research_os/plan.md
Browse files- phd_research_os/plan.md +23 -0
phd_research_os/plan.md
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# PhD Research OS — Current Plan
|
| 2 |
+
|
| 3 |
+
> Mutable task plan. Updated before and after every ECC lifecycle execution.
|
| 4 |
+
|
| 5 |
+
## Current Phase: Post-Initial Build
|
| 6 |
+
|
| 7 |
+
### Completed
|
| 8 |
+
- [x] Phase 0: Core data layer (22 tests passing)
|
| 9 |
+
- [x] Phase 1: Paper ingestion pipeline
|
| 10 |
+
- [x] Phase 2: Evaluation harness + regression gate
|
| 11 |
+
- [x] Phase 4: Obsidian export
|
| 12 |
+
- [x] Phase 5: Conflict detection
|
| 13 |
+
- [x] Phase 6: Hardening, backups, cost tracking
|
| 14 |
+
- [x] ECC Harness: Companion agent system (agent_os.py)
|
| 15 |
+
|
| 16 |
+
### Active
|
| 17 |
+
- [ ] Fine-tune model on GPU (train.py ready, needs GPU access)
|
| 18 |
+
- [ ] Process first batch of 5 papers through pipeline
|
| 19 |
+
- [ ] Create golden dataset (5 papers, manually annotated)
|
| 20 |
+
|
| 21 |
+
### Deferred
|
| 22 |
+
- [ ] Phase 3: ChromaDB vector search (needs embeddings from model)
|
| 23 |
+
- [ ] Phase 7-9: PhD Year 1 features (wait for real research data)
|