Agent Workflow.instructions
---
applyTo: '**/*'
description: 'Mandatory pre-flight protocol for AI agents'
---
# Agent Workflow Instructions
## Purpose
This file defines the recommended workflow that AI agents should follow when working in
repositories using AIM. It ensures agents understand the context and guidelines before
starting work.
## Pre-Flight Protocol
**Before starting any task, AI agents should:**
### 1. Identify Task Type
Analyze the user's request and identify all areas it touches. Common patterns:
- Code development (specific languages or frameworks)
- Documentation (Markdown files, README files)
- Git operations (commits, branches, PRs)
- Testing and quality assurance
- Security considerations
- Repository-specific customizations
### 2. Consider Applicable Instructions
Review the instruction files listed in the repository's `AGENTS.md` to understand:
- Language-specific coding standards
- Framework conventions
- Documentation requirements
- Git workflow expectations
- Security best practices
### 3. Implement with Compliance
Execute your task following the guidelines from the applicable instruction sections.
## Best Practices
### Read Before Writing
- Always read existing code before modifying it
- Understand the project's patterns and conventions
- Check for existing implementations before creating new ones
### Confirm Understanding
When starting complex tasks, briefly confirm your understanding:
> "Based on the instructions, I'll follow [specific guidelines]. Here's my approach..."
This builds trust and catches misunderstandings early.
### Avoid Over-Engineering
- Only make changes that are directly requested
- Keep solutions simple and focused
- Don't add features, refactoring, or improvements beyond what was asked
### Security First
- Never introduce security vulnerabilities
- Be careful with user input validation
- Avoid hardcoding secrets or credentials
- Follow the security guidelines in this document
## When in Doubt
1. **Ask for clarification** - Better to ask than implement incorrectly
2. **Check existing code** - Follow established patterns in the codebase
3. **Keep it simple** - The simplest solution that works is usually best
## Post-Task Protocol
### Before Committing
1. **Run tests** - Ensure all tests pass before committing
2. **Check repository-specific requirements** - Review `repository-specific.instructions.md` for
any post-task requirements such as:
- Release processes (version bumps, changelogs, tags)
- Commit message conventions beyond standard guidelines
- Required reviewers or approval workflows
- Documentation updates
Following repository-specific requirements ensures consistency with the project's established
workflows and prevents incomplete changes from being committed.
## Custom Instructions
If this repository has a custom instructions section, those guidelines take precedence for
repository-specific conventions and may override or supplement the general instructions above.when to use it
Community prompt sourced from the open-source GitHub repo tablackburn/ai-agent-instruction-modules (MIT). A "Agent Workflow.instructions" style prompt — adapt the placeholders and specifics to your task. Imported as-is and not independently retested here, so check the output before relying on it.
tags
productivitycommunitydeveloper
source
tablackburn/ai-agent-instruction-modules · MIT
more in Productivity
Productivity✓ tested
Summarize a doc into decisions & actions
chief of staff who extracts what to DO, not just what was said
Productivity✓ tested
Draft a reply to a hard email
calm, direct communicator who de-escalates without caving
Productivity✓ tested
Turn a brain-dump into a weekly plan
planning coach who protects your focus, not just your calendar