Project Overview

PromptPilot

SLM-powered control plane for AI coding agents.

PromptPilot uses a small-model harness around Codex/Claude-style agents to route prompts, clarify ambiguity, compress noisy context, preserve constraints, and decide when to pass through unchanged. The SLM manages the workflow; the frontier model writes and debugs the code.

PromptPilot optimizes for semantic-preserving context control, not blind token reduction.

First-time user? Start with QUICKSTART.md.

Docs strategy (lean repo + GitHub Wiki)

To keep this repository lightweight, long-form product/positioning documentation now lives in the GitHub Wiki.

• Wiki home: https://github.com/steyangdot/PromptPilot/wiki
• In-repo docs are kept minimal and operational (quickstart, security, contributing).
• Publish wiki pages from repo docs: scripts/publish_wiki.sh (details in docs/WIKI_WORKFLOW.md).

Install

Pick the extra that matches your coding agent:

pip install prpt[claude]      # for use with claude-code (Claude Haiku SLM)
pip install prpt[codex]       # for use with codex (GPT-5.4-nano SLM)
pip install prpt[all]         # both

[anthropic] / [openai] are kept as aliases for backward compatibility.