execution-plans

# Execution Plans Skill Treat execution plans as first-class artifacts, versioned in the repository. <!-- vendor-portability: declared. This skill reads and writes plan artifacts under the consumer's .agents/plans/ (active/, completed/, abandoned/) and .agents/debt/tech-debt-registry.md, seeded from .agents/plans/TEMPLATE.md. These are consumer-side write targets created on demand; a vendored install without .agents/ creates the directories when the first plan is written rather than failing silently. Issue #2050. --> ## Directory Structure ```text .agents/ ├── plans/ │ ├── active/ # Plans currently in progress │ ├── completed/ # Successfully finished plans │ └── abandoned/ # Plans that were stopped (with rationale) └── debt/ └── tech-debt-registry.md # Known technical debt items ``` ## Triggers | Trigger Phrase | Operation | |----------------|-----------| | `create execution plan` | Create new plan in active/ | | `update plan progress` | Add progress entry to existing plan | | `log decision` | Add decision to plan's decision log | | `complete plan` | Move plan to completed/ | | `abandon plan` | Move plan to abandoned/ with rationale | ## Plan Template Use `.agents/plans/TEMPLATE.md` as the starting point for new plans. ### Required Sections | Section | Purpose | |---------|---------| | Metadata | Status, dates, owner, complexity | | Objectives | Checkboxes for trackable goals | | Decision Log | Table of decisions with rationale | | Progress...