Prompt Details
Model
(gemini-3.8-flash)
Token size
269
Example input
[Insert Your Query or Task Here]: Analyze the following workload log and calculate total billable hours and non-billable hours: - Monday: 9:00 AM - 11:30 AM (Client Project A), 1:00 PM - 3:30 PM (Internal Admin) - Tuesday: 10:00 AM - 2:00 PM (Client Project B) - Wednesday: 8:30 AM - 12:00 PM (Client Project A), 1:00 PM - 2:30 PM (Team Training) - Thursday: 9:00 AM - 5:00 PM (Client Project B, with a 1-hour unpaid lunch break from 12:00 PM - 1:00 PM) Constraints: 1. Treat Client Projects A and B as billable; treat Admin and Training as non-billable. 2. Output the result strictly as a clean Markdown table with columns named "Category" and "Total Hours". 3. Do NOT use any bold formatting anywhere in the final output.
Example output
<scratchpad>
1. DECONSTRUCTION:
- Input: 4 days of time logs.
- Billable: Project A, Project B.
- Non-billable: Admin, Training.
- Constraints: Output ONLY a Markdown table (Category, Total Hours). NO bold formatting.
2. RETRIEVAL & BOUNDS:
- Math required for time durations.
- Monday:
A: 9:00 to 11:30 = 2.5 hrs
Admin: 13:00 to 15:30 = 2.5 hrs
- Tuesday:
B: 10:00 to 14:00 = 4.0 hrs
- Wednesday:
A: 8:30 to 12:00 = 3.5 hrs
Training: 13:00 to 14:30 = 1.5 hrs
- Thursday:
B: 9:00 to 17:00 (minus 1 hr lunch) = 7.0 hrs
3. DRAFTING (Calculations):
- Billable Total = 2.5 (A) + 4.0 (B) + 3.5 (A) + 7.0 (B) = 17.0 hrs
- Non-billable Total = 2.5 (Admin) + 1.5 (Training) = 4.0 hrs
Draft Table:
| Category | Total Hours |
| --- | --- |
| Billable | 17.0 |
| Non-billable | 4.0 |
4. HALLUCINATION & CONSTRAINT AUDIT:
- Math check: 2.5 + 4 + 3.5 + 7 = 17.0. (Correct). 2.5 + 1.5 = 4.0. (Correct).
- Bold formatting check: None in draft table.
- Extra text check: Just the table required? Yes, "strictly as a clean Markdown table". No extra conversational text allowed.
</scratchpad>
<final_answer>
| Category | Total Hours |
|---|---|
| Billable | 17.0 |
| Non-billable | 4.0 |
</final_answer>
Get this prompt + 9 more with SelectFrom $14/mo · Cancel anytimeFrom $14/mo · Cancel anytime
View Select
Instant access
Commercial use
Money‑backBy purchasing this prompt, you agree to our terms of service
5
Views
GEMINI-3.8-FLASH
Stop paying high API fees for flagship models. This system prompt forces low-cost models like Gemini 3.1 Flash-Lite to execute self-verification, audit their logic, and eliminate hallucinations before writing a word.
Features:
• Unlocks deep reasoning via <scratchpad> scaffolding
• Catches errors & bad syntax before final output
• Clean XML isolation for API parsing
• Built for Google AI Studio, n8n, Openclaw, Hermes, & custom stacks
...more
Added 3 weeks ago
