docs: vision-alignment check for task creation

docs/changelog/v0.0.7.md

@@ -36,6 +36,8 @@ Security hardening for running against untrusted target repositories.
 
 - **[meta]** Generate Work step (Step 6n) in evolve.md. After each session, the agent now scans the system across 7 dimensions (meta pipeline, code quality, repo health, architecture, agent DX, vision progress, security) and creates 1-5 follow-up tasks. Prevents the agent from being a passive task runner -- it actively identifies work. Enforces dimension diversity, duplicate checking, specific acceptance criteria, and a per-session cap of 5 tasks. Moves Meta-Prompt "Priority engine" from 0% to 50%. (task #0053)
 
+- **[meta]** Vision-alignment check for task creation. New `vision_section` frontmatter field on tasks (`loop1`, `loop2`, `self-maintaining`, `meta-prompt`, `none`). Before creating tasks, agents must check the last 5 tasks -- if 3+ target the same vision section, new tasks must prioritize a different section. Prevents tunnel vision where all work clusters in one area while others stagnate. Rule added to `docs/tasks/GUIDE.md` and enforced in evolve.md Step 6n constraints. Moves Meta-Prompt "Priority engine" from 50% to 60%. (task #0031)
+
 ## Fixed
 
 - **[cost]** Codex/OpenAI sessions now produce non-zero cost estimates. Added pricing for gpt-5.4 ($2.50/$15.00 per MTok), gpt-5.4-mini ($0.75/$4.50), and gpt-5.4-nano ($0.20/$1.25) to `MODEL_PRICING`. `parse_session_tokens()` now handles Codex `turn.completed` events (field mapping: `cached_input_tokens` -> cache_read, input adjusted to exclude cached). `record_session()` uses `AGENT_DEFAULT_MODELS` as model fallback when logs lack model identifiers. (`nightshift/constants.py`, `nightshift/costs.py`; task #0039)