diff --git a/docs/backups/settledreach.db.backup b/docs/backups/settledreach.db.backup
index a6fcdaba5..d79571e51 100644
Binary files a/docs/backups/settledreach.db.backup and b/docs/backups/settledreach.db.backup differ
diff --git a/docs/sprints/sprint-31/ci.md b/docs/sprints/sprint-31/ci.md
new file mode 100644
index 000000000..01bb4304d
--- /dev/null
+++ b/docs/sprints/sprint-31/ci.md
@@ -0,0 +1,57 @@
+# Sprint 31: Bedrock — CI Tasks
+
+**Goal:** Complete Phase 1 wiki content (cultural sweep, GTTR depth, star map popups) and refactor accumulated complexity in atlas.rs, main.gd, and Python tooling.
+
+**Branch:** `sprint-31/ci`
+**Agents:** Justine (build/deploy), Hoshe (QA)
+
+## New Tickets
+
+| # | Title | Blocked by |
+|---|-------|------------|
+| #777 | Add ruff linting + shared DB path module for Python scripts | — |
+
+Use `tooling/db/ticket show <id>` for full details.
+
+## Key Decisions
+
+- `decisions/process.md` — development process decisions
+
+## Notes
+
+### #777 ruff linting + shared DB path module (refactor)
+Python scripts in `tooling/db/` lack linting and share duplicated patterns.
+
+**Deliver:**
+1. **pyproject.toml** at repo root with ruff config (line length, target Python version, rule selection)
+2. **Shared module** `tooling/db/common.py` extracting:
+   - `resolve_db_path()` — uses `PROJECT_DB_PATH` env var with parent-dir fallback
+   - `load_config()` — reads `tooling/db/config.json`
+   - `get_connection()` — returns WAL-mode SQLite connection
+3. **Update all consumers** — `sqlite_connector.py`, `ticket`, `sprint`, `decisions_sync.py` to import from `common.py` instead of duplicating
+4. **Pre-push hook integration** — add `ruff check tooling/` to the pre-push hook alongside GDScript and Rust checks
+5. **Makefile target** — `make lint-python` running ruff
+
+**Key files:**
+- `tooling/db/sqlite_connector.py` — current DB path resolution
+- `tooling/db/ticket` — duplicated resolve pattern
+- `tooling/db/sprint` — duplicated resolve pattern
+- `tooling/db/decisions_sync.py` — duplicated resolve pattern
+- `tooling/db-backup` — bash equivalent
+
+**Verify:** `ruff check tooling/` passes. All DB CLI commands still work
+(`tooling/db/ticket count`, `tooling/db/sprint status`).
+
+## Dependency Chain
+
+```
+#777 (ruff + shared module) → standalone
+```
+
+## PR Workflow
+
+```bash
+tea pr create --repo jpmschweitzer/settled-reach --login schweitz \
+  --title "chore(ci): Sprint 31 — ruff linting + shared DB module" \
+  --description "body" --base main --head sprint-31/ci
+```
diff --git a/docs/sprints/sprint-31/client.md b/docs/sprints/sprint-31/client.md
new file mode 100644
index 000000000..53498610f
--- /dev/null
+++ b/docs/sprints/sprint-31/client.md
@@ -0,0 +1,73 @@
+# Sprint 31: Bedrock — Client Tasks
+
+**Goal:** Complete Phase 1 wiki content (cultural sweep, GTTR depth, star map popups) and refactor accumulated complexity in atlas.rs, main.gd, and Python tooling.
+
+**Branch:** `sprint-31/client`
+**Agents:** Stig (UI), Tyre (architecture), Hoshe (QA)
+
+## New Tickets
+
+| # | Title | Blocked by |
+|---|-------|------------|
+| #780 | Star map click-through — wiki/GTTR popup on system select | — |
+| #775 | Decompose main.gd and game_state.gd god objects | — |
+
+Use `tooling/db/ticket show <id>` for full details.
+
+## Key Decisions
+
+- `decisions/architecture.md` — D-013 (diegetic UI), D-020 (client architecture)
+- `decisions/scope.md` — development cascade, Phase 1 deliverable
+
+## Notes
+
+### #780 Star map click-through (Phase 1 deliverable)
+The star map insert module (#674, Sprint 30) renders 301 systems as a hop-ring
+view with click-to-select. Currently selecting a system updates `_selected_system`
+and shows the name in a basic info panel.
+
+**Deliver:** When a system is selected, show a popup/panel with:
+- System name, star type, hop distance, corridor
+- GTTR narrative excerpt (first paragraph of the wiki entry)
+- Body count, population summary
+- Adjacent systems (topology links)
+
+Data source: `client/data/star_map_data.json` already contains most fields.
+For GTTR text, either embed excerpts in the JSON (via `tooling/generate-star-map-data.py`)
+or load from a separate text file at runtime.
+
+**Key files:**
+- `client/ui/star_map.gd` — existing star map (525 lines)
+- `client/ui/star_map.tscn` — scene file
+- `client/data/star_map_data.json` — 4363 lines, 301 systems
+- `tooling/generate-star-map-data.py` — data generator (may need extending)
+
+### #775 Decompose main.gd and game_state.gd (refactor)
+Gemini code review flagged these as god objects:
+- `client/scripts/main.gd` (28KB) — UI management, input, state transitions
+- `client/scripts/autoloads/game_state.gd` (19KB) — global state, snapshots
+
+**Approach:** Extract focused components. Suggested splits:
+- `main.gd` → extract UI manager, insert state controller, input dispatcher
+- `game_state.gd` → extract snapshot handler, character state, session state
+
+**Constraint:** GDScript autoload parse-order rule (documented in CLAUDE.md).
+Autoloads can't reference `class_name` types at top level. Any new autoloads
+must follow the `load()` inline pattern. Prefer non-autoload scripts where possible.
+
+## Dependency Chain
+
+```
+#780 (star map popup) → standalone, Phase 1 priority
+#775 (decompose god objects) → standalone, refactor track
+```
+
+Independent — work in parallel.
+
+## PR Workflow
+
+```bash
+tea pr create --repo jpmschweitzer/settled-reach --login schweitz \
+  --title "feat(client): Sprint 31 — star map popup, main.gd decomposition" \
+  --description "body" --base main --head sprint-31/client
+```
diff --git a/docs/sprints/sprint-31/copy.md b/docs/sprints/sprint-31/copy.md
new file mode 100644
index 000000000..23edf9f9f
--- /dev/null
+++ b/docs/sprints/sprint-31/copy.md
@@ -0,0 +1,70 @@
+# Sprint 31: Bedrock — Copy Tasks
+
+**Goal:** Complete Phase 1 wiki content (cultural sweep, GTTR depth, star map popups) and refactor accumulated complexity in atlas.rs, main.gd, and Python tooling.
+
+**Branch:** `sprint-31/copy`
+**Agents:** Mellanie (author), Paula (narrative), Miri (worldbuilding)
+
+## New Tickets
+
+| # | Title | Blocked by |
+|---|-------|------------|
+| #766 | Cultural diversity sweep — wiki naming and heritage asymmetry | — |
+| #782 | Per-body GTTR entries for moons and stations | — |
+| #774 | Enrich GJ-672 (Nova Estrada) and GJ-147 (Entremeio) wiki entries | — |
+| #773 | Evaluate Iserlohn (GJ-532) planet name for IP originality | — |
+
+Use `tooling/db/ticket show <id>` for full details.
+
+## Key Decisions
+
+- `decisions/content.md` — D-NNN content decisions, cultural corridor system
+- `decisions/scope.md` — development cascade, Phase 1 deliverable definition
+
+## Notes
+
+### #766 Cultural diversity sweep (largest ticket)
+The cultural stereotyping audit (`docs/audits/cultural-stereotyping-audit.md`) flagged
+multiple cultural groups with imbalanced representation. Afrikaans is the worst at ~15:1
+confirming:subverting ratio (Sprint 30 added Brandpunt as first subversion). This ticket
+requires reviewing all cultural groups, identifying imbalances, and rewriting wiki entries
+to add subversions. Work across all 5 corridors. The audit's recommendations (tech hubs,
+academic centers, commercial dynasties, artistic communities) should guide the rewrites.
+
+**Scope rule:** This is a separate sprint's worth of work if done exhaustively. For this
+sprint, focus on the top 3 most imbalanced groups and deliver concrete wiki rewrites for
+at least 2 systems per group. File follow-up tickets for the remaining groups.
+
+### #782 Per-body GTTR entries
+Many systems have only system-level GTTR. Significant moons and stations (pop >10K) need
+their own entries. Existing patterns: `wiki/star-systems/GJ-*/gttr.md`. Prioritize hub
+systems at hops 1-4 first (highest traffic, most gameplay relevance).
+
+### #774 Thin wiki entries
+Paula's Sprint 30 review flagged Nova Estrada and Entremeio as noticeably thinner than
+the batch standard. Enrich to match Kettenschmied-level quality (economic logic + human
+cost in the same breath).
+
+### #773 Iserlohn IP evaluation
+Miri's review flagged "Iserlohn" (GJ-532 planet name) as a potential Legend of the Galactic
+Heroes reference. The real-German-city rationale is documented in the proposal notes. Make a
+deliberate decision: keep with rationale, or rename to another Sauerland-region reference.
+
+## Dependency Chain
+
+```
+#766 (cultural sweep) → standalone, largest piece
+#782 (per-body GTTR) → standalone, parallel track
+#774 (thin entries) → standalone, quick
+#773 (Iserlohn IP) → standalone, quick decision
+```
+
+All tickets are independent — work in parallel by agent.
+
+## PR Workflow
+
+```bash
+tea pr create --repo jpmschweitzer/settled-reach --login schweitz \
+  --title "data(copy): Sprint 31 — cultural sweep, GTTR depth, wiki quality" \
+  --description "body" --base main --head sprint-31/copy
+```
diff --git a/docs/sprints/sprint-31/joint.md b/docs/sprints/sprint-31/joint.md
new file mode 100644
index 000000000..5dc3d986e
--- /dev/null
+++ b/docs/sprints/sprint-31/joint.md
@@ -0,0 +1,45 @@
+# Sprint 31: Bedrock — Joint Notes
+
+**Goal:** Complete Phase 1 wiki content (cultural sweep, GTTR depth, star map popups) and refactor accumulated complexity in atlas.rs, main.gd, and Python tooling.
+
+## Sprint Completion Proof
+
+When Sprint 31 is done, you can:
+
+1. **Click any system** in the star map and see a popup with its name, star type, corridor, GTTR excerpt, body count, and adjacent systems
+2. **Read culturally diverse wiki entries** — at least 2 subversion entries per top-3 imbalanced cultural groups
+3. **Read GTTR entries for major moons and stations** — hub system bodies at hops 1-4 have individual entries
+4. **Run `ruff check tooling/`** and get a clean pass
+5. **Browse atlas.rs** and find subcommands in separate module files
+6. **Browse main.gd** and find it under 15KB with responsibilities extracted
+
+## Phase 1 Context
+
+This sprint advances Phase 1 of the development cascade (epic #746). Phase 1
+deliverable: "Implant-ready Godot map of the Reach with click-throughs + wiki/GTTR popups."
+
+**What's done:** 301/301 systems, body catalogs, star map insert, GTTR coverage
+**What this sprint adds:** Star map popups (#780), cultural quality (#766), GTTR depth (#782)
+**What remains after this sprint:** Heightmaps (#778), planetary screenshots (#779), economic profiles (#781) — these need design discussion before implementation, targeted for Sprint 32.
+
+## Not In Scope
+
+The following are **Phase 2-6 work** and must NOT be started until Phase 1 is complete:
+- Character creation production UI (#694)
+- Location-to-culture resolution (#679)
+- NPC personality surface area (#621)
+- Tycoon starting state design (#615)
+- Economic simulation
+- Settlement generation
+
+The cascade is enforced in the ticket system: #747 (Phase 2) is blocked by #746 (Phase 1).
+
+## Refactor Rationale
+
+The three refactors (#775, #776, #777) are infrastructure work, not feature development.
+They reduce complexity that compounds every sprint:
+- `atlas.rs` (2000+ lines) — every atlas feature adds to a single file
+- `main.gd` (28KB) + `game_state.gd` (19KB) — every client feature touches these
+- Python tooling (5 files with duplicated DB path logic) — every new script copies the pattern
+
+Doing these now prevents the debt from growing further as Phase 1 content work continues.
diff --git a/docs/sprints/sprint-31/server.md b/docs/sprints/sprint-31/server.md
new file mode 100644
index 000000000..181679ddc
--- /dev/null
+++ b/docs/sprints/sprint-31/server.md
@@ -0,0 +1,57 @@
+# Sprint 31: Bedrock — Server Tasks
+
+**Goal:** Complete Phase 1 wiki content (cultural sweep, GTTR depth, star map popups) and refactor accumulated complexity in atlas.rs, main.gd, and Python tooling.
+
+**Branch:** `sprint-31/server`
+**Agents:** Dudley (simulation), Tyre (architecture), Hoshe (QA)
+
+## New Tickets
+
+| # | Title | Blocked by |
+|---|-------|------------|
+| #776 | Split atlas.rs subcommands into modules | — |
+
+Use `tooling/db/ticket show <id>` for full details.
+
+## Key Decisions
+
+- `decisions/architecture.md` — D-010 (server-authoritative), D-020 (Rust/bevy_ecs)
+
+## Notes
+
+### #776 Split atlas.rs subcommands into modules (refactor)
+`server/src/bin/atlas.rs` is 2000+ lines containing all CLI subcommands in one
+file. Each subcommand (author, commit-system, corridor-status, list-systems,
+verify, next, show-system, etc.) should be extracted into its own module.
+
+**Approach:**
+- Create `server/src/bin/atlas/` module directory
+- Move `atlas.rs` → `server/src/bin/atlas/main.rs` (thin dispatch)
+- Extract each `cmd_*` function into its own file:
+  - `author.rs` — `cmd_author`, `generate_body_matrix`
+  - `commit.rs` — `cmd_commit_system`
+  - `corridor_status.rs` — `cmd_corridor_status`
+  - `list.rs` — `cmd_list_systems`, `cmd_next`, `cmd_unfinished`
+  - `show.rs` — `cmd_show_system`
+  - `verify.rs` — `cmd_verify`
+- Shared types/helpers stay in `main.rs` or a `common.rs`
+
+**Key file:** `server/src/bin/atlas.rs` (~2000 lines)
+
+**Verify:** `cargo build --bin atlas` must compile. `cargo test` must pass.
+Run `tooling/atlas stats` and `tooling/atlas corridor-status` to verify CLI
+still works after the split.
+
+## Dependency Chain
+
+```
+#776 (atlas split) → standalone
+```
+
+## PR Workflow
+
+```bash
+tea pr create --repo jpmschweitzer/settled-reach --login schweitz \
+  --title "refactor(simulation): Sprint 31 — split atlas.rs into modules" \
+  --description "body" --base main --head sprint-31/server
+```