agent-environment-retr…
Use when a completed session needs an agent-environment retrospective. Not for an engineering retrospective from telemetry: use…
Use when a user builds or styles a web frontend or asks to make it not look AI-generated. Not for component-level UI work without the token system: use frontend-ui-engineering.
Claude Code
$ npx -y skills add OutlineDriven/odin-claude-plugin --skill frontend-design-deslop --agent claude-code
Codex
$ npx skills add OutlineDriven/odin-claude-plugin --skill frontend-design-deslop -a codex -y
Cursor
$ npx skills add OutlineDriven/odin-claude-plugin --skill frontend-design-deslop -a cursor -y
Gemini CLI
$ npx skills add OutlineDriven/odin-claude-plugin --skill frontend-design-deslop -a gemini-cli -y
GitHub Copilot
$ npx skills add OutlineDriven/odin-claude-plugin --skill frontend-design-deslop -a github-copilot -y
OpenCode
$ npx skills add OutlineDriven/odin-claude-plugin --skill frontend-design-deslop -a opencode -y
Windsurf
$ npx skills add OutlineDriven/odin-claude-plugin --skill frontend-design-deslop -a windsurf -y
name: frontend-design-deslop description: 'Use when a user builds or styles a web frontend or asks to make it not look AI-generated. Not for component-level UI work without the token system: use frontend-ui-engineering.'
| Field | Bound contract | |---|---| | Trigger | User builds or styles a web frontend or asks to make it not look AI-generated. | | Authority | Reversible local: writes only DESIGN.md, design token files, and component CSS in the working project; rollback is version control. No remote mutation. No other files are touched. | | Side effect | Writes DESIGN.md, design tokens, and component CSS. | | Done | A committed token system and a crafted interface with a recorded slop-audit pass and a WCAG 2.2 AA pass/fail gate that passes. |
Required: the frontend project directory to style, and the surface to craft (page, component, or app shell).
Optional: brand palette, typeface preferences, existing token files, and a target framework. When omitted, derive a neutral token system from the project's existing styles; state that it was derived rather than supplied.
1. Read the target surface and any existing styles, tokens, and brand inputs. Record what was found versus supplied so the design system is grounded, not invented. Done when: found-versus-supplied is recorded. 2. Author a design strategy in DESIGN.md: layout grid, type scale, color system in OKLCH, spacing scale, motion intent, and component list. State the strategy before writing tokens so the tokens follow a decision, not a guess. Done when: DESIGN.md strategy section is written. 3. Emit a token system as CSS custom properties (or the project's token format) covering color, typography, spacing, radius, shadow, and motion. Tokens are the single source of truth; component CSS references tokens, never hard-coded values. Done when: token system covers all six categories and no component CSS uses hard-coded values. 4. Craft the interface using the tokens. Apply the never-slop list and reject every violation before claiming the surface done:
Done when: every never-slop item is checked and cleared or fixed. 5. Run a WCAG 2.2 AA pass/fail gate over the crafted surface: contrast ratios for text and UI components, focus visibility, target sizes, and semantic structure. Record pass or fail per criterion. Done when: every criterion is recorded as pass. 6. Run a slop audit against the never-slop list and record each item as cleared or violated. Fix violations in the interface and tokens before recording a pass. Done when: every item is recorded as cleared. 7. Commit DESIGN.md, the token files, and the component CSS. Record the slop-audit result and the WCAG 2.2 AA gate result alongside the commit so the pass is auditable. Done when: artifacts are committed with both gate results recorded.
Committed DESIGN.md, token system file, and component CSS implementing the crafted interface, plus a recorded slop-audit pass and WCAG 2.2 AA gate pass; the done predicate holds only when both gates pass and the artifacts are committed.
Formerly the ODIN Claude Plugin. The repository URL is unchanged. Outline-Driven Development, nicknamed ODIN, is a highly opinionated code-agent skill library: principles-first engineering, surgical editing, and workflow automation, published as installable
Use when a completed session needs an agent-environment retrospective. Not for an engineering retrospective from telemetry: use…
Use when a repo needs agent setup, AGENTS.md added or made lean, CLAUDE.md audited, or agent instructions scored or pruned. Not…
Use when a human explicitly asks for a full repository agent-compatibility pass returning a scored report with prioritized fixes.…
Use when setting up a project, auditing agent command permissions, or asking which read-only bash commands and domains to allow.…
Use when asked to build or review a CLI intended for coding agents and return flag-driven, pipeline-safe, idempotent design…
Use when the user asks to make the skills framework work in a new harness, IDE, or CLI. Not for remote, credential, publish,…
You MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or…
Ask which skill or flow fits your situation. A router over the skills in this repo.