diff --git a/.story_kit/stories/upcoming/60_status_based_directory_layout.md b/.story_kit/stories/upcoming/60_status_based_directory_layout.md
new file mode 100644
index 0000000..18c53b6
--- /dev/null
+++ b/.story_kit/stories/upcoming/60_status_based_directory_layout.md
@@ -0,0 +1,74 @@
+---
+name: Status-Based Directory Layout
+test_plan: pending
+---
+
+# Story 60: Status-Based Directory Layout
+
+## User Story
+
+As a developer, I want work items organized by pipeline status rather than by type, with a unified naming convention, so the directory structure reflects what stage everything is at.
+
+## Current Layout (mixed)
+
+```
+.story_kit/
+  current/
+  stories/
+    upcoming/
+    archived/
+  bugs/
+    archive/
+  spikes/
+  specs/
+```
+
+## New Layout
+
+```
+.story_kit/
+  upcoming/        ← all work items waiting to start
+  current/         ← being coded
+  qa/              ← being reviewed by QA agent
+  merge/           ← being merged to master
+  archived/        ← done
+  specs/           ← long-lived project info (not a work item)
+  worktrees/       ← agent worktrees (unchanged)
+  project.toml     ← config (unchanged)
+```
+
+## Naming Convention
+
+All work items use: `{number}_{type}_{slug}.md`
+
+```
+upcoming/57_story_foo_bar_blah.md
+upcoming/58_spike_a_b_c.md
+upcoming/59_bug_flappapa.md
+```
+
+Types: `story`, `bug`, `spike`. The number is the primary identifier, auto-incremented across all types. You say "story 57", "bug 59", etc.
+
+## Acceptance Criteria
+
+- [ ] Migrate existing `stories/upcoming/` → `upcoming/`, renaming files to include `_story_` type prefix
+- [ ] Migrate existing `stories/archived/` → `archived/`, renaming files to include `_story_` type prefix
+- [ ] Migrate existing `bugs/` → `archived/` (for closed) or `upcoming/` (for open), renaming to include `_bug_` type prefix
+- [ ] Remove old `stories/`, `bugs/` directories
+- [ ] `create_story`, `create_bug` MCP tools use the new naming convention and write to `upcoming/`
+- [ ] `start_agent` moves from `upcoming/` to `current/`
+- [ ] `accept_story` and `close_bug` move from `current/` (or `merge/`) to `archived/`
+- [ ] `find_story_file()` and all path references updated to search by status dirs
+- [ ] `next_story_number()` scans all status dirs for the highest number across all types
+- [ ] All existing tests updated for new paths
+- [ ] Integration test: full lifecycle through upcoming → current → archived
+
+## Supersedes
+
+- Story 51 (Deterministic Spike Lifecycle) — folded into this story's naming convention
+
+## Out of Scope
+
+- QA and merge pipeline automation (just create the empty directories)
+- Frontend changes to reflect new layout
+- Spike-specific MCP tools (create_spike, archive_spike) — follow-up