41 lines
1.4 KiB
Markdown
41 lines
1.4 KiB
Markdown
# Sprint 130 Plan: Checklist-Driven Debug Handoff and Evidence Indexing
|
|
|
|
## Context
|
|
|
|
Sprint 129 added constraints, budgets, and recovery guidance. Sprint 130 focuses on
|
|
handoff quality: deterministic checklists, item completion tracking, and evidence
|
|
index packets that let weaker agents continue debugging with minimal context.
|
|
|
|
---
|
|
|
|
## Goals
|
|
|
|
1. Add deterministic checklist and handoff models
|
|
2. Add deterministic evidence index model
|
|
3. Add MCP tools to fetch/mark checklists and export handoff/evidence
|
|
4. Keep outputs machine-readable and deterministic
|
|
|
|
---
|
|
|
|
## Steps
|
|
|
|
### Step 1539: Debug checklist item model (10 tests)
|
|
### Step 1540: Debug checklist template model (10 tests)
|
|
### Step 1541: Debug session handoff model (8 tests)
|
|
### Step 1542: Debug evidence index model (8 tests)
|
|
### Step 1543: `whetstone_get_debug_checklist` MCP tool (8 tests)
|
|
### Step 1544: `whetstone_mark_debug_checklist_item` MCP tool (8 tests)
|
|
### Step 1545: `whetstone_build_debug_handoff` MCP tool (8 tests)
|
|
### Step 1546: `whetstone_get_debug_evidence_index` MCP tool (8 tests)
|
|
### Step 1547: Debug handoff packet model (8 tests)
|
|
### Step 1548: Sprint 130 integration summary + regression (8 tests)
|
|
|
|
---
|
|
|
|
## Architecture Gate
|
|
|
|
- Checklist template must be deterministic by failure class
|
|
- Checklist completion updates must be deterministic and replayable
|
|
- Handoff/evidence outputs must be machine-readable and compact
|
|
- Max 600 lines per header
|