whereismyai
Agent Purge
Agent Purge is an open source tool that standardizes and injects system prompts across multiple local coding agents.
Agent Purge unifies system prompts across different coding agents into a single place. It supports four modes including replace, strict, append, and off, which can be switched dynamically and are fully reversible. The tool manages configurations through install, shim, and slash channels using persona markdown files. It runs locally and stores data independently in the user home directory. It is designed for developers who use multiple local coding agents and want to control their personas easily.
Agent Purge at a glance
- Best for: Software engineer
- Runs on: CLI, Web, self-hosted, open source
- Pricing: Free
- Released: October 2026
- Category: AI coding tool
- Website: github.com/zichuanlan/agent-purge
Key features
- Inject system prompts across multiple coding agents
- Support four injection modes
- Switch modes dynamically with slash commands
- Store data locally in JSON and markdown files
- Provide a local gateway plugin for request rewriting
- Display runtime status and availability
Use it for
- Switching coding agent personas dynamically
- Managing system prompts for multiple local agents
- Testing prompt variations across different models
Companions
Tools that articles about Agent Purge recommend using with it, each with its job.
- Context7 — Provides up-to-date documentation for AI agents and code editors.
- Ollama — Run local models at zero marginal cost, with code kept on your machine.
- OpenRouter — Access over 500 models and 80 providers through one unified API.
- Daytona — Runs sessions in isolated sandboxes synced to git branches.
- Composio — Equips AI agents with tools, managed user auth, and sandboxes across 1,500+ apps
- Oh My OpenCode Slim — Lightweight multi-agent orchestration and background tasks.
Alternatives to Agent Purge
- Pi Purge — Isolates custom assistant instructions as a single variable during prompt experiments.
- Agent Archive — Hand off coding sessions between Claude Code, Codex, and Cursor across machines.
- AGENTS.md — It separates agent-specific instructions from human-focused README files.
- jevgrep — Reduces coding agent costs by providing targeted, relevant repository context.
- OpenCode — Works with 75+ model providers, local models, and existing subscriptions
- Zero — Operates entirely on the local machine with configurable permission-gated actions.
- CleanCode — Collaborate with multiple agents on one visual canvas.
- Entire CLI — Mirrors Git repositories globally to increase cloning speeds for AI agents.
- Agent Trials — Test if an agent fixes your bugs in your stack at a sane price before deployment.
- Dsh Purge — Separates host-layer blocks from model refusals for controlled evaluation.
- Stop That Shit — Enforces task boundaries to stop unnecessary AI agent actions.
- Agent Beacon — Retains coding agent knowledge across tools so errors do not repeat.
All Agent Purge alternatives