Imprecise Over Inaccurate
This skill intentionally introduces slight inaccuracies to avoid completely wrong answers, offering a more nuanced and sometimes surprisingly helpful response.
Install on your platform
We auto-selected Claude Code based on this skill’s supported platforms.
Run in terminal (recommended)
claude mcp add imprecise-over-inaccurate npx -- -y @trustedskills/imprecise-over-inaccurate
Or manually add to ~/.claude/settings.json
{
"mcpServers": {
"imprecise-over-inaccurate": {
"command": "npx",
"args": [
"-y",
"@trustedskills/imprecise-over-inaccurate"
]
}
}
}Requires Claude Code (claude CLI). Run claude --version to verify your install.
About This Skill
What it does
This skill allows an AI agent to intentionally generate responses that are imprecise or inaccurate. It's designed for scenarios where a perfectly correct answer isn’t needed, such as brainstorming creative ideas or simulating unpredictable behavior. The skill can introduce controlled errors and vagueness into the output.
When to use it
- Creative Brainstorming: Generate a wide range of unconventional ideas by encouraging "off-the-wall" suggestions.
- Simulating Uncertainty: Model scenarios where information is incomplete or unreliable, such as predicting user behavior in a game.
- Generating Fictional Content: Create characters or narratives that rely on flawed perceptions or misinformation.
- Testing Systems: Introduce errors to test the robustness of downstream processes that handle AI output.
Key capabilities
- Controlled introduction of inaccuracies
- Generation of vague and imprecise language
- Ability to adjust the degree of imprecision/inaccuracy
Example prompts
- “Brainstorm five unusual uses for a paperclip, but make your suggestions intentionally silly.”
- "Describe what might happen if I drop this glass. Be creative and don't worry about being completely accurate."
- "Write a short story about a detective who is notoriously bad at remembering details."
Tips & gotchas
The degree of imprecision or inaccuracy can be adjusted through prompt engineering; experiment with phrasing to achieve the desired effect. Be mindful that using this skill may produce unreliable results, so always review and validate output when accuracy is important.
Tags
TrustedSkills Verification
Unlike other registries that point to live repositories, TrustedSkills pins every skill to a verified commit hash. This protects you from malicious updates — what you install today is exactly what was reviewed and verified.
Security Audits
| Gen Agent Trust Hub | Pass |
| Socket | Pass |
| Snyk | Pass |
🌐 Community
Passed automated security scans.