spike · openclaw — Claude Code skill: what it does, how to install
spike runs throwaway prototypes to test technical feasibility, compare alternate approaches, and deliver an evidence-based verdict before building.
MIT 296 installs ★ 395,001 updated 2026-10-25
npx skills add openclaw/openclaw --skill spike -a claude-codeWhat spike does
The skill handles structured technical spikes. It states the core question, looks into relevant docs, builds the smallest runnable artifact, and exercises an edge case. After running the test, it returns an explicit verdict: VALIDATED, PARTIAL, or INVALIDATED, complete with reproduction commands and recommendations.
Files live in .tmp/openclaw-spikes/<slug> by default, or inside spikes/<NNN-slug>/ with a local README.md when tracked code is requested. It prefers isolated CLI tools, single endpoints, or bare HTML snippets over heavy frameworks, container images, or environment setup files.
Spike code must never be merged directly into a production codebase without a full rewrite. It serves only to answer questions early, ruling out dead ends before real implementation begins.
What it helps with
- test feasibility with runnable code before committing resources
- compare competing approaches on equal terms and shared metrics
- inspect external dependencies for maintenance activity and licensing
- document evidence for abandoning an unworkable architecture path
Example prompts for the agent
- Spike this library to see if it parses our legacy data format
- Build a quick prototype comparing approach A and B for this endpoint
- Is this background sync pattern possible before we build the feature?
How to install spike
1Skills CLI (Vercel)
npx skills add openclaw/openclaw --skill spike -a claude-code2Manually: put the files in place
~/.claude/skills/spike/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/spike/
1Skills CLI (Vercel)
npx skills add openclaw/openclaw --skill spike -a codex2Manually: put the files in place
.agents/skills/spike/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add openclaw/openclaw --skill spike -a cursor2Manually: put the files in place
.cursor/skills/spike/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install openclaw/openclaw skills/spike2Manually: put the files in place
.github/skills/spike/SKILL.md1Skills CLI (Vercel)
npx skills add openclaw/openclaw --skill spike -a gemini-cli2Platform CLI
gemini skills install https://github.com/openclaw/openclaw --path skills/spikeManually: put the files in place
.devin/skills/spike/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add openclaw/openclaw --skill spike -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about spike
How do I install spike in Claude Code?
Run npx skills add openclaw/openclaw --skill spike -a claude-code. Alternatively, place the definition at ~/.claude/skills/spike/SKILL.md.
When should spike not be used?
Skip it if the question is easily answered by checking docs or reading existing code, or whenever production-ready implementation is requested upfront.
Where does the skill write prototype files?
It creates files in .tmp/openclaw-spikes/<slug> by default, or under spikes/<NNN-slug>/ with a README.md if repository tracking is explicitly requested.
Source text
SKILL.md · MIT
---
name: spike
description: Run throwaway prototypes to validate feasibility, compare approaches, and report a verdict.
metadata: { "openclaw": { "emoji": "🧪" } }
---
# Spike
Use when the user wants to test an idea before committing to a real build: "spike this", "quick prototype", "is this possible", "compare A/B", "before we build".
Do not use when reading docs/code can answer the question, or when the user clearly asked for production implementation.
Loop
1. Question: state the concrete feasibility question.
2. Research: read enough docs/source to choose credible approach.
3. Build: create the smallest runnable artifact that validates or invalidates the idea.
4. Stress: try one edge case or failure mode.
5. Verdict: `VALIDATED`, `PARTIAL`, or `INVALIDATED`.
Output shape
- Default workspace: `.tmp/openclaw-spikes/<slug>` unless user asks for a tracked repo-local path.
- Repo-local option: `spikes/<NNN-slug>/` with `README.md` and minimal code.
- Prefer runnable CLI, tiny HTML, one endpoint, or focused test.
- Avoid package sprawl, Docker, env files, app frameworks, and production cleanup.
Multi-spike ideas
- Split into 2-5 independent questions.
- Run the riskiest question first.
- For A/B comparisons, keep inputs equal and measure the same dimensions.
- Ask before building all variants if the work is more than a small prototype.
Verdict format
```markdown
## Verdict: VALIDATED | PARTIAL | INVALIDATED
Question: ...
Evidence: exact command/output/measurement.
What worked: ...
What failed or surprised us: ...
Recommendation: ship / adjust / avoid, with the next production step.
```
Rules
- An invalidated spike is useful when it rules out a path with evidence.
- Do not merge spike code into production without rewriting it normally.
- If external dependencies are evaluated, check health: recent release/commit, docs, license, install friction.
© openclaw · MIT · openclaw/openclaw
Similar skills
All skills for Agent setup: memory, context, orchestration →spike by other authors comes first.