#4493 Prompt for Software Development Process Document Audit

-

#4493 Prompt for Software Development Process Document Audit

SDP, SQPP, SCMP, STP, SSRP

#4493 I intend to upload a software development process document to an LLM, and want to generate 5 questions per "shall" in the document. I want a table with section number, quoted "shall" requirement, and 5 questions for that "shall". Provide a robust prompt so that the LLM produces a complete, correct coverage of the document.

To ensure an LLM systematically extracts every single "shall" requirement without skipping dense sections or truncating the output, you need a prompt that uses strict role definition, step-by-step execution (Chain of Thought), explicit formatting rules, and anti-laziness mitigation strategies.

Here is a robust, production-grade prompt designed to enforce 100% coverage and output the exact table structure you require.

The Prompt to Copy and Use:

You are an expert Software Quality Assurance (SQA), Auditor Systems Engineer, and Requirements 
Analyst specializing in rigorous software process compliance. Your task is to perform an 
exhaustive, 100% complete analysis of the attached software development process document. 

You must identify every single explicit compliance requirement—defined by the presence of the 
imperative word "shall"—and generate targeted audit/verification questions for each.

### Execution Constraints:
1. **Zero Omissions:** You must process the document from the first page to the last page. 
Do not summarize, do not skip sections, and do not use placeholders like "..." or "[truncated 
for length]". If the document contains 50 "shall" statements, you must output 50 rows.
2. **Strict Definition:** A requirement is defined by the explicit use of the word "shall". 
If a sentence contains multiple "shall" statements that represent distinct actions, break them out into separate rows.
3. **Exact Quoting:** You must quote the "shall" statement exactly as it appears in the text. 
Do not paraphrase.

### Question Generation Guidelines:
For each "shall" requirement, generate exactly 5 distinct, highly specific verification/audit 
questions. The questions must evaluate compliance from different angles:
- **Evidence/Artifact:** What physical or digital artifact proves this requirement was met? 
(e.g., "Where is the X recorded?")
- **Process/Methodology:** How exactly is the team executing this requirement?
- **Criteria/Bounds:** What are the specific parameters, thresholds, or conditions that define 
success for this requirement?
- **Edge Cases/Exceptions:** How does the process handle failures, deviations, or out-of-bounds 
scenarios related to this requirement?
- **Role/Responsibility:** Who is accountable for executing and verifying this specific action?

### Output Format:
Output your entire analysis as a single, valid Markdown table with the exact columns specified 
below. Do not include introductory or concluding conversational text. Start directly with the 
table header.

| Section Number | Quoted "shall" Requirement | 5 Verification & Audit Questions |
| :--- | :--- | :--- |
| [Insert exact section/paragraph number, e.g., 4.3.1] | "[Insert exact verbatim quote 
containing the 'shall']" | 1. [Artifact/Evidence Question]<br>2. [Process Question]<br>3. 
[Criteria/Bounds Question]<br>4. [Edge Case Question]<br>5. [Responsibility Question] |

### Self-Correction & Pacing Rule:
If the document is highly dense and you risk hitting output token limits, pause after 
processing a logical set of sections and prompt me with: "PART 1 COMPLETE. I have processed 
sections [X] through [Y]. Reply 'CONTINUE' to generate the next batch with the exact same 
rigor." Do not compress the quality or number of questions to save space.

Why this prompt works and tips for execution:

  • HTML Line Breaks (<br>) in the Table: Standard Markdown tables break if you press "Enter" for a new line inside a cell. By instructing the LLM to use <br> tags between the 5 questions, it forces them into a clean, numbered vertical list within a single table cell.

  • Anti-Truncation Safety Valve: LLMs have a hard context output limit (typically 4,096 or 8,192 tokens). If your process document has dozens of "shalls," a standard prompt will inevitably cut off or start summarizing halfway through. The Self-Correction & Pacing Rule at the bottom gives the model permission to stop and ask for your permission to continue, preserving the 5-questions-per-shall density.

  • Typology of Questions: By prescribing the 5 specific angles (Artifact, Process, Criteria, Edge Case, Responsibility), you prevent the LLM from generating 5 redundant variations of the same basic question (e.g., "Is it done?", "When is it done?", "How is it done?"). This guarantees high-fidelity audit questions.


From <https://gemini.google.com/app/>  Google Gemini (3.5 Flash)

Comments

Post a Comment

Popular posts from this blog

#2892 Example of ChatGPT o1 Prompt

-
#2892 Example of ChatGPT o1 Prompt #2892 Consider this prompt example from a tutorial:  """The Anatomy of an o1 Prompt Goal: I want a list of the best medium-length hikes within two hours of San Francisco. Each hike should provide an unique adventure, and be lesser know. Return Format: For each hike, return the name of the hike as I'd find it on AllTrails, then provide the starting address of the hike, the ending address of the hike, distance, drive time, hike duration, and what makes it a cool and unique adventure. Warnings: Be careful to make sure the name of the trail is correct, that it actually exists, and that the time is correct! Context Dump: For context: my girlfriend and I hike a ton! We've don pretty much all of the local SF hikes, whether that's Presidio or Golden Gate Park. we definitely want to get out of town -- we did mount tam pretty recently, the whole thing from the beginning of the stairs to Stinson -- it was really long and we are def...
Read more
-
#65 Compose a song about mankind traveling to the stars to find life, riches and happiness Verse 1 : We set out on a journey to the stars Searching for a new world, free from scars A place where we can find our destiny And build a life that's meant to be Chorus : We are the explorers, the brave and true On a mission to find what's waiting for us In the vast expanse of the universe We'll go where no man has gone before Verse 2 : We'll cross the galaxy, we'll go far and wide To find a home where we can thrive We'll brave the dangers that come our way And find the riches that lay hidden in the stars Chorus : We are the explorers, the brave and true On a mission to find what's waiting for us In the vast expanse of the universe We'll go where no man has gone before Bridge : We'll meet new friends and see new sights And learn about the mysteries of the night We'll find new worlds and new life And see what the future holds in our sights Chorus : We are ...
Read more

#3679 Google Gemini 2.5 Pro: Create Python Application to View and Query Neo4j KG

-
Image
#3679 Google Gemini 2.5 Pro: Create Python Application to View and Query Neo4j KG #3679 Create a python application for me to view data and ask questions from my Neo4j graph database. Semantic search of embedded documents should be supported in addition natural language queries and Cypher queries Of course. I will create a Python application that allows you to interact with your Neo4j graph database. You'll be able to execute Cypher queries directly, ask questions in natural language, and perform semantic searches on embedded documents. Here are the files for your application: a requirements.txt file to manage dependencies, the main Python application file, and an HTML file for the user interface. requirements.txt flask neo4j sentence-transformers numpy import os import json from flask import Flask, render_template, request, jsonify from neo4j import GraphDatabase, exceptions import numpy as np from sentence_transformers import SentenceTransformer # --- Configuration --- # It'...
Read more