skills

KO | EN

gitlite — search

#graphql #gpt #devops #privacy #codex #openai #database #go #svg #hacktoberfest

(★ 9)

AI agent skills — build once, personalize anywhere. Install with: openclaw skills install <slug>

File Explorer

ZIP Download

- SKILL.md
- SKILL.md
- - blindspot-pass.md
  - design-directions.md
  - mock-before-wire.md
  - options-brainstorm.md
  - reference-semantics-map.md
  - spec-interview.md
  - teach-me-explainer.md
  - tweakable-plan.md
- SKILL.md
- SKILL.md
- SKILL.md
- - HANDOFF_TEMPLATE.md
- SKILL.md
- design-inspiration-v0.1.0.zip
- ecosystem-mapping-v0.1.0.zip
- finding-unknowns-v0.1.0.zip
- graph-max-v0.1.0.zip
- handoff-spec-v0.1.0.zip
- problem-mining-v0.1.0.zip
- workspace-os-v0.11.1.zip
- workspace-os-v0.12.0.zip
- - desire-mapping.md
  - interview-flow.md
  - root-cause-reframing.md
- SKILL.md
- - examples.md
- SKILL.md
- - claude-plugin.json.tmpl
  - CLAUDE.md.tmpl
  - codex-plugin.json.tmpl
  - compound-learn-SKILL.md.tmpl
  - dashboard.html.tmpl
  - doctor-SKILL.md.tmpl
  - doctor.py.tmpl
  - janitor-schedule.tmpl
  - janitor-SKILL.md.tmpl
  - marketplace.json.tmpl
  - process-input-SKILL.md.tmpl
  - reflect-SKILL.md.tmpl
  - routine.md.tmpl
  - skillify-SKILL.md.tmpl
  - work-AGENTS.md.tmpl
  - work-package-README.md.tmpl
  - work-plan.md.tmpl
  - work-README.md.tmpl
- - improvements-ledger.md
  - README.md
  - rubric.md
  - test_work_continuity.py
  - validate.py
- - blueprint.md
  - domain-inference.md
  - packaging.md
  - self-management.md
  - skill-authoring.md
  - skillifying-work.md
- SKILL.md
.gitignore
package.sh
project-overview.md
README.md

# Use via CDN

jsDelivr

jsDelivr serves any public GitHub repository as a CDN with zero setup. Pick a version and a file to get a ready-to-paste link and snippet.

Version

File

Link

Example

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

AI agent skills — build once, run anywhere.

Built by [@sidneyswift](https://x.com/sidneyswift).

## Install

```bash
# OpenClaw
openclaw skills install <skill-name>

# Claude Code / Codex / any agent
npx skills add sidneyswift/skills
```

## Skills

| Skill | Description |
| --- | --- |
| [design-inspiration](./design-inspiration/) | Break through design blocks by systematically mining nine curated high-signal galleries — social posts, web UI, minimal sites, OG images, decks, logos, branding, editorial, and brand guidelines. Structured sweep → cross-pollinate → extract patterns → restart momentum. |
| [ecosystem-mapping](./ecosystem-mapping/) | Take a core solution and map every friction point around it into a compounding product ecosystem. Walks through core transformation extraction, friction mapping (before/during/after), two-sided market scans, and reinforcement mapping — so products strengthen each other instead of standing alone. |
| [finding-unknowns](./finding-unknowns/) | Surface gaps between the task description and what the work actually requires — before writing code. Quadrant-based discovery with techniques like blindspot passes, design directions, mocks, spec interviews, and tweakable plans. |
| [spec-goals](./spec-goals/) | Turn a durable objective into an evidence-based, ready-to-paste Codex `/goal`, or identify when the work needs discovery or a normal prompt instead. |
| [gauntlet-loop](./gauntlet-loop/) | Run iterative build-critique loops that push AI output past "good enough" toward a concrete quality bar. Decompose the goal, assign separate builders and critics, compare against a real reference, and keep looping until the output wins. For games, websites, writing, code, design, or anything where output can be inspected and improved. |
| [graph-max](./graph-max/) | Turn any hand-drawn workflow graph into a working multi-agent script. Draw the graph (whiteboard, paper, Excalidraw), send the image to a coding agent with one prompt, get a running orchestration script back. Teaches graph literacy — nodes, edges, fan-outs, feedback loops, decision gates. |
| [handoff-spec](./handoff-spec/) | Produce a handoff document for a stateless successor agent to pick up in-progress work when the current agent is low on usage credits, context, or switching platforms. Combines a stable Goal/Spec section with a point-in-time Handoff Status section — done, out-of-scope, remaining, nice-to-haves, live in production, merged but not deployed, still in progress — verified against real repo/PR/issue state rather than conversation memory. |
| [problem-mining](./problem-mining/) | Interview-driven discovery that surfaces profitable problems hiding in your own experience. 1-question-at-a-time interview across four phases — achievements, problems overcome, desire mapping, and root-cause reframing — then synthesizes validated product ideas. |
| [workspace-os](./workspace-os/) | Build a self-managing workspace OS with resumable work packages, shared agent instructions, domain skills, an installable plugin, report-only health checks, and evidence-backed learning loops. |

## License

MIT

skills

File Explorer

# Use via CDN

Was this content helpful?

Recent Feedback

Download README