savepoint 1.0.1 → 1.0.3

package/.savepoint/releases/v1.1/epics/E15-hardening/tasks/T003-debug-flag.md

@@ -0,0 +1,30 @@
+---
+id: E15-hardening/T003-debug-flag
+status: planned
+objective: Add --debug flag and SAVEPOINT_DEBUG env var for structured debug logging
+depends_on: []
+---
+
+# T003: Add --debug Flag and SAVEPOINT_DEBUG Env Var
+
+## Context Files
+
+- `main.go` — CLI entrypoint, --version flag
+- `cmd/` — command dispatch
+- `internal/board/` — board init, file watchers, command dispatch
+
+## Acceptance Criteria
+
+- [ ] --debug flag accepted at CLI level
+- [ ] SAVEPOINT_DEBUG env var recognized
+- [ ] Debug output includes board init, file watcher events, and command dispatch
+- [ ] Debug output is off by default (no performance impact)
+- [ ] `go test ./...` passes with no regressions
+
+## Implementation Plan
+
+- [ ] Define debug logging helper in a shared location
+- [ ] Add --debug flag parsing in main.go
+- [ ] Instrument key points in board init, file watcher, and update dispatch
+- [ ] Add tests for debug flag behavior
+- [ ] Run `make build && make test`

package/.savepoint/releases/v1.1/epics/E15-hardening/tasks/T004-dist-checksums.md

@@ -0,0 +1,27 @@
+---
+id: E15-hardening/T004-dist-checksums
+status: planned
+objective: Generate checksums.txt during make dist using Go crypto APIs
+depends_on: []
+---
+
+# T004: Add Distribution Checksums
+
+## Context Files
+
+- `internal/buildtool/main.go` — dist() function that creates tar.gz archives
+- `Makefile` — dist target
+
+## Acceptance Criteria
+
+- [ ] dist() generates checksums.txt file with SHA256 hashes
+- [ ] checksums.txt is included in the dist directory
+- [ ] Existing archive creation behavior preserved
+- [ ] `go test ./...` passes
+
+## Implementation Plan
+
+- [ ] Add SHA256 checksum computation in dist() using crypto/sha256
+- [ ] Write checksums.txt to dist directory
+- [ ] Add tests for checksum generation
+- [ ] Run `make build && make test`

package/.savepoint/releases/v1.1/epics/E15-hardening/tasks/T005-windows-targets.md

@@ -0,0 +1,28 @@
+---
+id: E15-hardening/T005-windows-targets
+status: planned
+objective: Add Windows amd64 and arm64 build targets to buildtool
+depends_on: []
+---
+
+# T005: Add Windows Build Targets
+
+## Context Files
+
+- `internal/buildtool/main.go` — targets list, build logic, localExecutable
+
+## Acceptance Criteria
+
+- [ ] Windows amd64 target added
+- [ ] Windows arm64 target added
+- [ ] .exe suffix handled for Windows binaries
+- [ ] Existing Linux and Darwin targets preserved
+- [ ] `go test ./...` passes
+
+## Implementation Plan
+
+- [ ] Add windows-amd64 and windows-arm64 to targets list
+- [ ] Handle .exe suffix in build output path
+- [ ] Update localExecutable to detect Windows
+- [ ] Update tests for new targets
+- [ ] Run `make build && make test`

package/.savepoint/releases/v1.1/epics/E15-hardening/tasks/T006-abbreviation-splitting.md

@@ -0,0 +1,26 @@
+---
+id: E15-hardening/T006-abbreviation-splitting
+status: planned
+objective: Fix splitChecklistSentences to skip periods preceded by known abbreviations
+depends_on: []
+---
+
+# T006: Fix Abbreviation Handling in Checklist Sentence Splitting
+
+## Context Files
+
+- `internal/board/detail.go:88` — splitChecklistSentences function
+
+## Acceptance Criteria
+
+- [ ] Known abbreviations (e.g., "e.g.", "i.e.") do not trigger sentence breaks
+- [ ] Existing sentence splitting behavior preserved for non-abbreviation periods
+- [ ] Abbreviation list is configurable/extensible
+- [ ] `go test ./...` passes with no regressions
+
+## Implementation Plan
+
+- [ ] Define known abbreviation set
+- [ ] Add abbreviation check before period-based sentence split
+- [ ] Add test cases for abbreviations in sentences
+- [ ] Run `make build && make test`

package/.savepoint/releases/v1.1/epics/E15-hardening/tasks/T007-root-test-allowlist.md

@@ -0,0 +1,28 @@
+---
+id: E15-hardening/T007-root-test-allowlist
+status: planned
+objective: Move agent_skills_test.go to a test package and extract audit allowlist to a named constant
+depends_on: []
+---
+
+# T007: Move Root-Level Test and Extract Audit Allowlist
+
+## Context Files
+
+- `agent_skills_test.go` — root-level package main test
+- `internal/board/epic_panel.go:116-119` — allowedSections map
+
+## Acceptance Criteria
+
+- [ ] agent_skills_test.go moved to cmd_test package (or appropriate location)
+- [ ] allowedSections extracted to a named constant with documentation
+- [ ] All existing tests still pass after refactoring
+- [ ] `go test ./...` passes with no regressions
+
+## Implementation Plan
+
+- [ ] Move agent_skills_test.go to an appropriate internal test package
+- [ ] Update imports and paths in moved test
+- [ ] Extract allowedSections map to a named constant
+- [ ] Update any references in epic_panel.go
+- [ ] Run `make build && make test`

package/.savepoint/releases/v1.1/epics/_archived/T001-cli-entrypoint.md

@@ -0,0 +1,25 @@
+---
+id: E06-audit-command/T001-cli-entrypoint
+status: in_progress
+objective: "Create CLI entrypoint for savepoint audit command"
+depends_on: []
+---
+
+# T001: CLI Entrypoint
+
+## Acceptance Criteria
+
+- `savepoint audit --help` shows usage: `audit <epic-id|release> [--skip --reason]`
+- `savepoint audit <epic-id>` runs audit for that epic
+- `savepoint audit <release>` runs audit for all unaudited epics in that release
+- `savepoint audit --skip --reason "..."` skips the audit and logs reason
+- Unknown arguments show error and exit non-zero
+
+## Implementation Plan
+
+- [x] Add `cmd/audit.go` with CLI argument parsing
+- [x] Wire audit command into main.go
+- [x] Implement arg validation for epic-id (E##) and release (v#.#) formats
+- [x] Add subcommand dispatch to audit workflow
+- [x] Test `savepoint audit --help` output
+- [x] Run `make build && make test`

package/.savepoint/releases/v1.1/epics/_archived/T002-quality-gates.md

@@ -0,0 +1,27 @@
+---
+id: E06-audit-command/T002-quality-gates
+status: done
+objective: "Implement quality gate runner that blocks on failure"
+depends_on: ["E06-audit-command/T001-cli-entrypoint"]
+phase: build
+---
+
+# T002: Quality Gates
+
+## Acceptance Criteria
+
+- Reads `quality_gates` from `.savepoint/config.yml`
+- Runs each configured command (lint, typecheck, test)
+- Blocks audit if any gate fails and `block_on_failure: true`
+- Continues if `block_on_failure: false` but logs failure
+- Shows clear output: which gates passed/failed
+
+## Implementation Plan
+
+- [x] Add `internal/audit/gates.go`
+- [x] Read config `quality_gates` struct
+- [x] Implement `RunGates(ctx) error` that executes each command
+- [x] Handle null gates (skip if not configured)
+- [x] Add exit code propagation for blocking behavior
+- [x] Test with configured gates in .savepoint/config.yml
+- [x] Run `make build && make test`

package/.savepoint/releases/v1.1/epics/_archived/T003-snapshot.md

@@ -0,0 +1,27 @@
+---
+id: E06-audit-command/T003-snapshot
+status: planned
+objective: "Generate file tree and changed-files snapshot"
+depends_on: ["E06-audit-command/T002-quality-gates"]
+---
+
+# T003: Snapshot Generation
+
+## Acceptance Criteria
+
+- Generates `.savepoint/audit/{release}/{epic}/snapshot.md`
+- Includes file tree (gitignore-respecting)
+- Includes list of changed files (from git)
+- Does NOT include file contents (no code in snapshot)
+- Snapshot path matches expected format
+
+## Implementation Plan
+
+- [ ] Add `internal/audit/snapshot.go`
+- [ ] Implement `GenerateSnapshot(epic, release) error`
+- [ ] Walk directory tree respecting .gitignore
+- [ ] Run `git diff --name-only` for changed files
+- [ ] Write snapshot.md with tree + changed-files sections
+- [ ] Create audit directory if not exists
+- [ ] Test snapshot output format
+- [ ] Run `make build && make test`

package/.savepoint/releases/v1.1/epics/_archived/T004-ai-reconcile.md

@@ -0,0 +1,29 @@
+---
+id: E06-audit-command/T004-ai-reconcile
+status: planned
+objective: "Generate AI prompt for semantic review, parse findings"
+depends_on: ["E06-audit-command/T003-snapshot"]
+---
+
+# T004: AI Reconcile (Semantic Review)
+
+## Acceptance Criteria
+
+- Generates a prompt file for human to give to their AI
+- Prompt includes: snapshot, epic E##-Detail.md, changed files list, AGENTS.md Code Style rules
+- Prompt instructs AI to check against 10 Code Style rules
+- After human runs AI, parses findings for "Must Fix Before Close" section
+- Fails audit if "Must Fix Before Close" has any items
+
+## Implementation Plan
+
+- [ ] Add `internal/audit/reconcile.go`
+- [ ] Implement `GeneratePrompt(epic, release) (string, error)`
+- [ ] Embed 10 Code Style rules in prompt template
+- [ ] Include snapshot.md content in prompt
+- [ ] Include epic E##-Detail.md in prompt
+- [ ] Include changed files list in prompt
+- [ ] Implement `ParseFindings(output) (hasIssues bool, err)`
+- [ ] Check for "Must Fix Before Close" header and non-empty content
+- [ ] Test prompt generation and finding parsing
+- [ ] Run `make build && make test`

package/.savepoint/releases/v1.1/epics/_archived/T006-tui-review.md

@@ -0,0 +1,31 @@
+---
+id: E06-audit-command/T006-tui-review
+status: planned
+objective: "Implement TUI review mode for approving/rejecting proposals"
+depends_on: ["E06-audit-command/T005-proposals"]
+---
+
+# T006: TUI Review Mode
+
+## Acceptance Criteria
+
+- Shows each proposal as a side-by-side diff (old vs new)
+- User can Approve, Reject, or Edit each proposal
+- Edit opens inline editor to modify the "With" section
+- High-divergence warning if proposal changes >50% of file
+- Shows progress: X of Y proposals reviewed
+- Keyboard: arrow nav, Enter approve, Backspace reject, e edit, q quit
+
+## Implementation Plan
+
+- [ ] Add `internal/audit/review.go` with Bubble Tea model
+- [ ] Create `ReviewModel` with proposals, current index, decisions
+- [ ] Implement `RenderProposal` showing old/new side-by-side
+- [ ] Add keyboard handlers: Up/Down (nav), Enter (approve), Backspace (reject), e (edit)
+- [ ] Implement edit mode: text area for "With" content
+- [ ] Add high-divergence check (config threshold from config.yml)
+- [ ] Show progress bar: "Proposal 1 of 5"
+- [ ] Track decisions: approved/rejected/edited map
+- [ ] Return decisions to caller for apply step
+- [ ] Test review flow manually
+- [ ] Run `make build && make test`

package/.savepoint/releases/v1.1/epics/_archived/T008-skip-handling.md

@@ -0,0 +1,34 @@
+---
+id: E06-audit-command/T008-skip-handling
+status: planned
+objective: "Implement audit skip with logging"
+depends_on: ["E06-audit-command/T001-cli-entrypoint"]
+---
+
+# T008: Skip Handling
+
+## Acceptance Criteria
+
+- `--skip --reason "..."` flags are required together
+- Missing reason shows error
+- Skipped audit logged to `.savepoint/audit-log.md`
+- Log entry includes: timestamp, epic, release, reason, skipped-by
+- Skipped epic still advances router (no changes to docs)
+
+## Implementation Plan
+
+- [ ] Add `internal/audit/log.go`
+- [ ] Implement `LogSkip(epic, release, reason) error`
+- [ ] Parse existing audit-log.md or create new
+- [ ] Append new entry in format:
+    ```markdown
+    ## {timestamp}
+
+    - **Epic:** {epic}
+    - **Release:** {release}
+    - **Reason:** {user reason}
+    - **Status:** ⚠ skipped
+    ```
+- [ ] Update router to advance past skipped epic
+- [ ] Test skip flow with valid/invalid args
+- [ ] Run `make build && make test`

package/.savepoint/releases/v1.1/v1.1-PRD.md

@@ -0,0 +1,139 @@
+---
+version: 1.1
+name: "Board Polish & Navigation"
+status: in_progress
+---
+
+# Release v1.1 — Board Polish & Navigation
+
+## Overview
+
+v1.1 is the "love what you look at" release. It takes the functional board from v1 and makes it polished, reliable, and pleasant to navigate. The four epics cover layout robustness, multi-platform tooling, visual refinement, and a new epic navigation paradigm.
+
+**v1 delivered:** a working Kanban board with phase transitions.
+**v1.1 delivers:** a board that doesn't flicker on resize, builds everywhere, looks intentional, and treats epics as first-class navigable objects.
+
+## What ships in v1.1
+
+1. **TUI Optimisation (E01)** ✅ *Audited and complete*
+   - Next Activity indicator in header
+   - Naming conventions reconciled (Design.md → E##-Detail.md, PRD.md → v1-PRD.md)
+   - Virtual viewport scrolling for columns and detail overlay
+   - Column focus border stability — no content shift when switching columns
+   - All 7 tasks done
+
+2. **Cross-Platform Compatibility (E02)** — *Not started*
+   - `make build`, `test`, `clean` work on Windows, Linux, macOS
+   - Linux amd64 + arm64 and macOS amd64 + arm64 build targets
+   - Smoke tests and versioned release artifacts
+   - 4 tasks planned
+
+3. **UI Visual Refinement (E03)** — *Not started*
+   - Right-border clipping fixed on terminal resize
+   - Next Activity as a dedicated line below header with phase-aligned styling
+   - Checkbox rendering at sentence boundaries
+   - Width arithmetic audit — content fills available space exactly
+   - 4 tasks planned
+
+4. **Epic Navigation (E04)** — *Not started*
+   - Epic sidebar becomes a focusable, navigable component
+   - ↑/↓ to navigate epics, ←/→ to enter/exit sidebar
+   - Enter opens an "Epic Detail" overlay showing E##-Detail.md content
+   - Existing `E` key dropdown retained
+   - 2 tasks planned
+
+5. **Tasking Permissions & Router Updates (E05)** ✅ *Audited and complete*
+   - Agents can only set `status: in_progress`; user alone sets `done` or retreats
+   - New TUI `p` priority hotkey explicitly updates router to the focused non-done task
+   - Router task priority decoupled from navigation — browse without side effects
+   - Priority writes keep router in `task-building`; audit handoff remains explicit
+   - All 6 tasks done
+
+6. **Audit Command (E06)** — *Planned*
+   - `savepoint audit <epic-id|release>` runs full pipeline
+   - Quality gates block on failure
+   - AI semantic review checks 10 Code Style rules — blocks if "Must Fix Before Close" has items
+   - TUI review mode for approving/rejecting proposals
+   - Updates Design.md last_audited, marks epic status: audited
+   - 9 tasks planned
+
+7. **Init Command (E07)** — *Planned*
+   - `savepoint init [dir] [--force] [--install]` scaffolds .savepoint/
+   - Target validation (empty/compatible/already-initialized)
+   - Template-based scaffold with project name interpolation
+   - Magic prompt output to stdout + clipboard
+   - Best-effort clipboard (doesn't fail on error)
+   - 7 tasks planned
+
+8. **Board Command (E08)** — *Planned*
+   - `savepoint board [--release <release>] [--epic <epic>]` launches TUI
+   - `savepoint` (no args) defaults to board
+   - Bubble Tea TUI with 3-column Kanban
+   - Detail pane for selected task
+   - Keyboard navigation (arrows, vim-style j/k/h/l)
+   - Status transitions with gate enforcement
+   - Non-TTY plain table fallback
+   - 8 tasks planned
+
+9. **Doctor Command (E09)** — *Planned*
+   - `savepoint doctor [--epic <epic>]` runs diagnostics
+   - Config, router, structure, YAML validation
+   - Dependency checks (missing deps, cycles, duplicates)
+   - Audit state and orphan detection
+   - Ad-hoc quality gate runner
+   - Human-readable diagnostic output with repair suggestions
+   - Exit codes: 0=clean, 1=problems, 2=error
+   - 6 tasks planned
+
+10. **Task File Validation & Auto-Fix (E12)** — *In Progress*
+     - Default `phase: build` when `status=in_progress` but phase missing
+     - Default `status: planned` when both status and column missing
+     - Better error hints with suggested fixes
+     - Validate on write with helpful messages
+
+11. **Codebase Audit Remediation (E13)** — *In Progress*
+    - Fix cycle detection bug producing inaccurate error paths
+    - Remove dead code, stdlib reimplementations, hardcoded state maps
+    - Centralize duplicated logic (normalization, frontmatter extraction, shared utilities)
+    - Fix AtomicWrite cross-device fallback and config accent defaults
+    - Add quality gate timeout
+    - Remove committed binaries, add gitignore and linter config
+    - Decompose `update.go` monolith into focused handlers
+    - Extract synchronous I/O from `Update()` into `tea.Cmd` async pattern
+    - Add test coverage for untested packages
+    - 7 tasks planned
+
+## Epic breakdown
+
+| # | Epic | Status | Tasks |
+|---|------|--------|-------|
+| 01 | TUI Optimisation | Audited | 7/7 done |
+| 02 | Cross-Platform Compatibility | Audited | 4/4 done |
+| 03 | UI Visual Refinement | Planned | 0/4 done |
+| 04 | Epic Navigation | Audited | 2/2 done |
+| 05 | Tasking Permissions & Router Updates | Audited | 6/6 done |
+| 06 | Audit Command | Planned | 0/9 done |
+| 07 | Init Command | Planned | 0/7 done |
+| 08 | Board Command | Planned | 0/8 done |
+| 09 | Doctor Command | Planned | 0/6 done |
+| 10 | Task File Validation & Auto-Fix | In Progress | 5/5 done |
+| 13 | Codebase Audit Remediation | In Progress | 0/7 done |
+
+## Success criteria
+
+- `make build` and `make test` pass on Windows, Linux, and macOS
+- Terminal resize at any width ≥ 40 produces no visual corruption
+- Content fills the available terminal width exactly at every breakpoint
+- Epic sidebar is navigable with ↑/↓, and Enter shows a detail overlay
+- All existing v1 functionality is preserved unchanged
+- `go build ./...` and `go test ./...` pass
+
+## Known issues
+
+- E03 has `Design.md` instead of the expected `E03-Detail.md` — rename to match convention (tracked in E01/T002 scope, may need a follow-up task)
+
+## Risks
+
+- **Build tooling scope creep.** E02 is pure Makefile work, but smoke-testing on non-Windows requires either CI or manual verification on actual Linux/macOS.
+- **Naming convention holdout.** E03's `Design.md` may cause discovery or rendering issues if the board ever reads `E##-Detail.md` by pattern. Low priority — handled as a rename task if needed.
+- **Epic panel is ≥120 only.** E04's features require a wide terminal — users on narrow terminals still use the `E` key dropdown.

package/.savepoint/router.md

@@ -1,136 +1,57 @@
 # Agent State Machine
 
-This file routes the agent based on the project's current state. Read this whenever you start a session.
+## Read order
 
-## Read order on every session
+1. This file (router.md)
+2. Current state → next action
+3. Active epic E##-Detail.md
+4. Active task file
 
-1. This file (you are here)
-2. The current state below to know what to do next
-3. The active epic Design
-4. The active task file, when a task is selected
-
-Read `.savepoint/PRD.md` only for project vision changes. Read `.savepoint/Design.md` only for architecture changes or audit closeout. Read `.savepoint/releases/v1/PRD.md` only for release planning or epic-order questions.
-
-**Conditional read (token discipline):** if your active task touches **TUI rendering, theme, or visual design**, also read `.savepoint/visual-identity.md` after Design.md. Otherwise skip it — it's ~1.8K tokens you don't need.
+Read `.savepoint/PRD.md` only for vision changes. Read `.savepoint/Design.md` only for architecture/audit.
 
 ## Current state
 
 ```yaml
 state: task-building
 release: v1.1
-epic: E01-tui-optimisation
-task: E01-tui-optimisation/T001-border-resize-fix
-next_action: Build v1.1 E01 T001 border resize fix.
+epic: E14
+task: E15-hardening/T001-benchmarks
+next_action: Build E15-hardening/T001-benchmarks.
 ```
 
-## State → next action
-
-<!-- AGENT: Read the state above. Find the matching block below. Follow it. -->
-
-### `state: pre-implementation`
-
-The project has its PRD and Design locked but no epics defined yet.
-
-**Next action:**
-
-1. Read `.savepoint/releases/v1/PRD.md` — the v1 release scope (epic list lives there).
-2. Help the user define the epics list and confirm priority.
-3. For each epic in order, create the directory `.savepoint/releases/v1/epics/E##-{epic-name}/` with a `Design.md` stub.
-4. When epic E01 (scaffolding) is created, transition to `state: epic-design` for that epic.
-
-**Do not** start writing code. We are still in planning.
-
-### `state: epic-design`
-
-An epic exists but its `Design.md` is empty or a stub.
-
-**Next action:** Walk the user through filling out the epic's `Design.md`:
-
-- What is this epic adding to the system?
-- What components / files does it touch?
-- What's the architectural delta vs the current state?
-
-When complete, transition to `state: epic-task-breakdown` for this epic.
-
-### `state: epic-task-breakdown`
-
-Epic Design exists but tasks are missing or not fully planned.
-
-**Next action:**
+## State → action
 
-1. Re-read the epic Design.
-2. Create or update the full epic task list — each task **independently buildable**, **objective-led**, with declared `depends_on`.
-3. Each task file lives at `.savepoint/releases/v1/epics/{E##-epic}/tasks/TNNN-slug.md` with frontmatter:
-   ```yaml
-   ---
-   id: {E##-epic}/TNNN-slug
-   status: planned
-   objective: "<one sentence>"
-   depends_on: []
-   ---
-   ```
-4. In the same pass, write each task's `## Implementation Plan` as inline `- [ ]` checkboxes.
-5. When every task is planned, transition to `state: task-building` for the first unblocked task.
+### pre-implementation
 
-### `state: task-planning`
+PRD + Design locked, no epics yet.
 
-Reserved for repair or late-added tasks. Normal epic planning happens during `state: epic-task-breakdown`.
+**Next:** 1) Read release PRD for epic list, 2) Define + confirm epic order, 3) Create epic stubs. Transition to `epic-design` for E01.
 
-**Next action:** Read the task's `objective`. Write the implementation plan as inline `- [ ]` checkboxes under a `## Implementation Plan` heading. Set `status: planned`. Stop.
+### epic-design
 
-### `state: task-building`
+Epic E##-Detail.md is empty/stub.
 
-Task is `in_progress`. All `depends_on` are `done`.
+**Next:** Define what this epic adds, files it touches, and architectural delta. Then transition to `epic-task-breakdown`.
 
-**Next action:** Execute the plan. Tick checkboxes as you complete them. Edit code per the **Code Style** rules in `AGENTS.md`. When all checkboxes tick, set `status: review` and stop.
+### epic-task-breakdown
 
-### `state: audit-pending`
+Epic exists, tasks missing.
 
-The last task in an epic is `done`. Audit must run before the next epic starts.
+**Next:** 1) Re-read epic, 2) Create task files at `tasks/TNNN-slug.md` with `status: planned`, `objective`, `depends_on`, 3) Add `## Implementation Plan` checkboxes per task. When all planned → first unblocked task.
 
-**Next action:** Confirm `.savepoint/audit/{E##-epic}/snapshot.md` exists. If it is missing while the audit CLI is still unavailable, create one manual snapshot from the known epic scope once; do not search broadly for replacement inputs. Then read the snapshot, read the epic's `Design.md`, and read only the files listed as changed. Write one patch-shaped proposal bundle to `.savepoint/audit/{E##-epic}/proposals.md`:
-
-- `Design.md` section — merge only the epic delta into project architecture.
-- `AGENTS.md` section — refresh Codebase Map entries from changed-module metadata; preserve existing rows.
-- `epic-Design.md` section — add "implemented as:" notes and deltas from the original plan.
-- `Quality Review` section — semantic-review findings against the 10 Code Style rules (advisory only).
-
-Prefer delta-only edits (`Insert After`, `Replace`, `Delete`) anchored to exact text. Do not quote and replace entire large sections unless the whole section genuinely changed.
-
-Proposal format:
-
-```md
-## Target File
-
-`path/to/file.md`
-
-## Replace
-
-<exact old heading, marker, or section>
-
-## With
-
-<new content>
-```
-
-Quality review section format:
-
-```md
-## Must Fix Before Close
-
-## Carry Forward
-
-## Already Fixed
-```
+### task-building
 
-After proposals are approved, apply approved proposals to live files, mark the epic `Design.md` as `status: audited`, update project `Design.md` `last_audited`, refresh `AGENTS.md` Codebase Map, and advance this router to the next epic state.
+Task `in_progress`, depends satisfied.
 
-Stop. The user reviews proposals in the TUI before commit actions.
+**Next:** When starting work, set task `status: in_progress` and press `p` in TUI to mark the focused task as router priority. Execute plan, tick checkboxes, run quality gates, update router to next task or `audit-pending`. Stop.
 
-## Capability check
+### audit-pending
 
-If you are not Claude Opus / Gemini 2.5 Pro / GPT-5.5 / equivalent, surface a warning to the user:
+Epic complete, needs audit before next epic.
 
-> _"Heads up — I'm running on a lighter model. Savepoint's planning steps work best with top-tier models because the embedded prompts are detailed. I'll do my best, but consider switching the model for PRD/Design/Task-breakdown steps."_
+**Next:** Fresh audit agent reads epic E##-Detail.md, task files, drift notes, Design.md, AGENTS.md, and scoped changed files. Write one audit file to `.savepoint/releases/{release}/epics/{E##-epic}/E##-Audit.md`:
+- `## Main Findings` user-facing narrative only
+- `## Code Style Review` checklist against AGENTS.md rules
+- `## Proposed Changes` admin/apply blocks using `### Target File`, `### Replace`, `### With`
 
-Then proceed.
+After user approves: apply proposals, mark epic `status: audited`, update Design.md `last_audited`, advance router.