Menu
Use before planning when requirements are ambiguous, large, user-visible, or likely to cause rework unless decisions are locked.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Use only when the user explicitly allows bounded unattended execution of an approved Alloy task.
Use when encountering a bug, regression, failing command, or unexpected behavior before attempting a fix.
Use when implementing an approved Alloy plan. Maintains progress.md, enforces TDD/debug discipline, and stops on blocked gates.
Use when a task needs broad source understanding before planning or execution.
Use when turning context or a clear user request into an implementation-ready plan with acceptance criteria, files, commands, gates, risks, and handoff.
Automated end-to-end QA — read acceptance criteria (Azure DevOps or manual), generate Playwright test plan, execute with screenshots/video/GIF, produce acceptance report. Uses Playwright CLI for deterministic, reportable execution.
| name | alloy-discuss |
| description | Use before planning when requirements are ambiguous, large, user-visible, or likely to cause rework unless decisions are locked. |
Use this skill to turn fuzzy user intent into decision-locked context for Planner and Builder.
Write or update:
.alloy/tasks/<task-id>/context.md
Use a stable task id from the user, issue id, branch name, or a short slug. Do not create multiple ids for the same request.
# <task-id>: Context
## User Request
- Original request:
- Follow-up corrections:
- Explicit constraints:
## Decisions Locked
- Decision:
- Chosen:
- Reason:
- Rejected alternatives:
## Acceptance Signals
- User-visible success:
- Non-goals:
- Compatibility expectations:
## Open Questions
- [ ] Question:
- Why it blocks planning:
- Options:
## Source Reality To Check In /plan
- Files/directories:
- Commands:
- Docs/specs:
## Handoff To /plan
- Task id:
- Planning focus:
- Risks:
Ask at most one blocking question at a time in chat. Prefer discovering codebase facts yourself. Ask the user only about intent, tradeoffs, and externally unknowable decisions.
Planner may spawn one-shot Explorer only for broad source discovery. Do not create persistent subagent files.
Return: