Use when creating a Spec Kit technical plan, research notes, data model, contracts, quickstart, or validation plan from an approved spec.
Scanned 9/5/2026
Install to Claude Code
npx -y skills add me-cedric/agent-compass --skill speckit-plan --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Speckit Plan?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/me-cedric-speckit-plan)More formats (shields.io, HTML) on the badges page.
---
name: speckit-plan
description: >
Use when creating a Spec Kit technical plan, research notes, data model,
contracts, quickstart, or validation plan from an approved spec.
risk_level: medium
writes_files: true
requires_tools: []
---
# Spec Kit Plan
Read `AGENTS.md`, the approved `spec.md`, `.specify/memory/constitution.md` if
present, and existing architecture/module docs.
## Do
1. Convert behavior into technical approach, affected files, data model,
contracts, risks, and validation commands.
2. Put implementation decisions here, not in the spec.
3. Resolve unknowns in `research.md` before tasks are generated.
4. Keep API contracts, env vars, docs, security, migrations, and shared package
consumers in the validation map when affected.
## Done
- `plan.md` is executable and traceable to the spec.
- Research/open questions are resolved or marked blocking.
- Validation commands come from the host command registry or package scripts.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!
Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...