Repo hygiene: version sync, changelog, archive stale docs, README update

- Bump pyproject.toml to 0.1.2 (was still 0.1.0)
- Add v0.1.1 and v0.1.2 changelog entries
- Archive stale CORE-manual.md and FMT-SR-manual.md to docs/internal/archive
  (described old imports, deps, and architecture that no longer match the repo)
- Remove empty src/__init__.py (src/she/ is the canonical package)
- Add "Repo status" table to README (version, canonical path, maturity, stable
  vs experimental surface)
- Update README examples table with notebooks
- Tighten worked-example narrative: explicitly describe what each score is and
  that these are heuristics, not topological invariants

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: Mircus

CHANGELOG.md

@@ -1,5 +1,52 @@
 # Changelog
 
+## v0.1.2 — Export, Decay Windows, Real Data (2026-03-26)
+
+### Added
+
+- **Result export** — `ranked_items_to_csv/json`, `bridges_to_csv/json`, `cohesion_to_csv/json`
+  for feeding SHE outputs into pandas, dashboards, or downstream tools
+- **Decay-weighted temporal accumulation** — `decay_window()` applies exponential
+  decay (configurable half-life) so older interactions fade rather than vanish at
+  a hard boundary
+- **Real public dataset notebook** — `examples/eu_email_analysis.ipynb` analyses the
+  SNAP EU Email Dept3 network (20 most-active members, 812 interactions, 27 months)
+  with Louvain community detection, temporal bridge/cohesion tracking, and matplotlib
+  plots
+- **Bundled dataset** — `data/eu_email_dept3.jsonl` (pre-processed from SNAP)
+
+### Tests
+
+- 41 tests passing (10 new: 6 export, 4 decay window)
+
+---
+
+## v0.1.1 — Modeling Layer, Social Analysis, Temporal (2026-03)
+
+### Added
+
+- **`SHEHyperstructure`** — decorated, weighted higher-order relational structure
+  with entity attributes, typed relations, and arbitrary per-simplex metadata
+- **Social analysis** — `rank_diffusers`, `rank_entity_diffusers`,
+  `rank_simplex_diffusers`, `find_bridge_simplices`, `group_cohesion`,
+  `rank_influencers` (graph centrality vs simplex diffusion contrast)
+- **Bulk ingestion** — `from_records`, `from_csv`, `from_jsonl`
+- **Temporal windowing** — `window()` and `rolling_windows()` for time-sliced analysis
+- Worked use case: `examples/social_media_diffusers.py`
+- Temporal notebook: `examples/temporal_diffusion_analysis.ipynb`
+
+### Fixed
+
+- Diffusion scoring: switched to rank-percentile normalisation with `lambda=1.0`
+  (previously scores saturated at 1.0 due to identity-term dominance)
+- `analyze_diffusion` no longer truncates to top 10 internally
+
+### Tests
+
+- 31 tests across 7 files (up from 8 in v0.1.0)
+
+---
+
 ## v0.1.0 — Research Preview (2025)
 
 First public release of SHE as a small, runnable research preview.

README.md

@@ -14,6 +14,17 @@ diffusion analysis**.
 This is a **Research Preview** released under the non-commercial
 [HNCL v1.0](LICENCE.md) license (not OSI open-source).
 
+## Repo status
+
+| | |
+|---|---|
+| **Version** | 0.1.2 |
+| **Canonical package** | `src/she/` — install with `pip install -e .`, import as `import she` |
+| **Maturity** | Research preview — API may change between releases |
+| **Stable surface** | `SHEHyperstructure`, social analysis functions, temporal windowing, export |
+| **Experimental** | `src/core/` (legacy monolith, convolution), `src/morse/` (discrete Morse) — not part of the public API |
+| **Tests** | 41 passing (CI on Python 3.10–3.12) |
+
 ## Why SHE?
 
 Standard graph analysis collapses every interaction to a pairwise edge.
@@ -34,23 +45,30 @@ Use SHE when:
 SHE does not replace graph libraries.  It adds a layer for the cases where
 graphs are not enough.
 
-## What v0.1 includes
+## What v0.1.x includes
 
 **Modeling layer**
 - `SHEHyperstructure` — decorated, weighted higher-order relational object
-  with entity attributes, typed relations, and bulk record ingestion
+  with entity attributes, typed relations, and bulk ingestion (`from_csv`, `from_jsonl`)
 
 **Social analysis**
 - `rank_diffusers` / `rank_entity_diffusers` / `rank_simplex_diffusers`
 - `find_bridge_simplices` — cross-community higher-order bridges
 - `group_cohesion` — structural cohesion scoring for candidate groups
 - `rank_influencers` — graph centrality vs. simplex diffusion comparison
 
+**Temporal**
+- `window` / `rolling_windows` — hard time-range slicing
+- `decay_window` — exponential decay with configurable half-life
+
+**Export**
+- `ranked_items_to_csv/json`, `bridges_to_csv/json`, `cohesion_to_csv/json`
+
 **Core simplicial engine**
 - Simplicial-complex construction (wraps [TopoNetX](https://github.com/pyt-team/TopoNetX))
 - Graph-to-simplicial lifting via clique detection
 - Hodge-Laplacian spectral analysis and harmonic-form extraction
-- Diffusion centrality ranking
+- Diffusion centrality ranking (rank-percentile normalisation)
 - Minimal matplotlib visualisation
 
 ## Installation
@@ -88,27 +106,35 @@ for b in find_bridge_simplices(hs):
 
 ## Worked use case: social-media diffusers
 
-`examples/social_media_diffusers.py` builds a two-community social scenario
+`examples/social_media_diffusers.py` builds a synthetic two-community scenario
 where a high-degree hub dominates graph centrality, but a cross-community
-triad is the actual diffusion engine.  The example compares graph-only
-ranking with simplex-level analysis and shows where they disagree.
+triad carries more weight as a higher-order structure.
 
 ```bash
 python examples/social_media_diffusers.py
 ```
 
 Output highlights:
-- **Graph centrality** ranks the hub (u0) first.
-- **Bridge detection** surfaces the {u3, u5, u7} triad as the top
-  cross-community structure.
-- **Group cohesion** scores the triad as structurally tight despite
+- **Graph centrality** (eigenvector on the 1-skeleton) ranks the hub (u0) first —
+  it has the most and heaviest pairwise edges.
+- **Bridge detection** (heuristic: community-span x relation weight) surfaces the
+  {u3, u5, u7} triad as the top cross-community structure — it spans both
+  communities and carries high group weight.
+- **Group cohesion** (geometric mean of internal weight, pair density, and
+  higher-order support) scores the triad as structurally tight despite
   containing no individually prominent member.
 
+These are heuristic scores, not topological invariants. The point is that
+graph-only centrality never even sees group-level structures, while SHE makes
+them queryable.
+
 ## Examples
 
 | Script | Description |
 |--------|-------------|
 | `examples/social_media_diffusers.py` | Graph vs. simplex ranking on a two-community scenario |
+| `examples/eu_email_analysis.ipynb` | Real-data notebook: SNAP EU Email network, temporal bridge/cohesion plots |
+| `examples/temporal_diffusion_analysis.ipynb` | Synthetic temporal scenario: bridge formation over three periods |
 | `examples/toy_triangle.py` | Smallest nontrivial complex — Hodge Laplacian printout |
 | `examples/social_group_lift.py` | Lift a small social graph to simplices via cliques |
 | `examples/group_diffusion_demo.py` | Weighted Karate Club diffusion analysis |

docs/CORE-manual.md docs/internal/archive/CORE-manual.mddocs/CORE-manual.md renamed to docs/internal/archive/CORE-manual.md

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "she"
-version = "0.1.0"
+version = "0.1.2"
 description = "Simplicial Hyperstructure Engine -- weighted simplicial-complex toolkit for higher-order relational modelling and Hodge diffusion analysis."
 authors = [{name = "Holomathics"}]
 readme = "README.md"