BitterPillEngineering

Audit AI assistant instruction sets and classify over-prompting rules into categorized outputs.

1|Updated Apr 16, 2026

Search Tags:#Claude Code #settings.json #token-efficiency #prompt-auditing #instruction-trimming #cross-file-conflicts #prompt-hygiene

Ready to run

npx skills add https://github.com/davdunc/pai-framework --skill bitterpillengineering-davdunc

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: BitterPillEngineering
Source: https://github.com/davdunc/pai-framework/tree/main/skills/BitterPillEngineering
Command: npx skills add https://github.com/davdunc/pai-framework --skill bitterpillengineering-davdunc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

BitterPillEngineering reduces over-prompting by auditing instruction sets and identifying rules that add scaffolding, ambiguity, redundancy, or unnecessary complexity that can degrade model performance.

Core Features & Use Cases

Over-prompting audits: Scans force-loaded files and evaluates each rule using a “smarter model” test plus Five Questions to classify what to cut, resolve, merge, evaluate, sharpen, move, or keep.
Full and fast workflows: Runs a comprehensive Audit workflow across all loaded instruction sources, or a QuickCheck workflow for a single file/rule block.
Actionable reporting: Produces a categorized report with estimated line/token savings to help prioritize trimming without breaking anti-fragile requirements.

Quick Start

Ask your AI to run the BitterPillEngineering Audit workflow on your Claude setup to generate a classified over-prompting report with estimated token savings.

Frequently Asked Questions about BitterPillEngineering

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I audit my Claude Code instruction files for over-prompting?▼

To audit Claude Code instruction files for over-prompting, you can run a comprehensive workflow that evaluates each rule against a smarter-model test and five diagnostics to classify rules for cutting, merging, or keeping.

What causes over-prompting in settings.json and how does it affect model outputs?▼

Over-prompting in settings.json occurs when instruction rules add scaffolding, ambiguity, or redundancy that degrades model performance. Auditing identifies these unnecessary complexities and estimates token savings for sharper outputs.

Can I check a single rule block for redundancy without scanning all loaded instruction sources?▼

Yes, you can run a QuickCheck workflow to perform a targeted fast evaluation of a single instruction file or rule block, applying the same smarter-model test without scanning the entire setup.

How do I resolve cross-file conflicts in my AI assistant instruction set?▼

To resolve cross-file conflicts in an AI assistant instruction set, an audit reads settings.json to detect loaded instruction sources and classifies conflicting rules into RESOLVE or MERGE categories for actionable trimming.

What is the smarter-model test for prompt hygiene and instruction trimming?▼

The smarter-model test for prompt hygiene evaluates whether a specific instruction rule is necessary by checking if a more capable model would still need that scaffolding, helping classify rules to CUT or SHARPEN.

Does prompt auditing work for large AI setups with multiple force-loaded files?▼

Yes, prompt auditing supports large AI setups by reading settings.json to identify all force-loaded files, detecting cross-file conflicts, and generating a categorized report with estimated line and token savings.