thuong180702/IdeaGauntlet
Stress-test a product idea file with adversarial agents and produce a verdict report — for use in CI / on pull requests.
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
Inputs
| name | description | required | default |
|---|---|---|---|
| idea-file | Path to the idea file to critique (Markdown or text). | no | IDEA.md |
| mode | Analysis mode: 'court' (deep multi-role debate) or 'quick' (fast critique, fewer tokens). | no | court |
| api-key | LLM API key. Anthropic (sk-ant-...) or Groq (gsk_...). Pass a repo secret — never hardcode. | yes | — |
| model | Optional model override (e.g. claude-haiku-4-5-20251001 to save tokens). | no | "" |
| version | idea-gauntlet npm version or dist-tag to run. | no | latest |
| output-file | Where to write the generated Markdown report. | no | idea-gauntlet-report.md |
Outputs
| name | description |
|---|---|
| report-file | Path to the generated Markdown report (empty if the idea file was missing). |
| skipped | 'true' when the idea file did not exist and nothing ran. |