AI Project Management Glossary: Key Terminology and Concepts
This glossary provides quick reference for all terminology, concepts, and frameworks covered in the Claude for Project Managers guide. Use it for review, reference, and teaching others.
Core Concepts
Project Brain
A persistent collection of context documents that give Claude deep understanding of your specific project. Components include the Context Readme, Stakeholder Register, Decision Log, and Risk Register. The Project Brain eliminates the need to re-explain your project in every conversation.
Context Window
The amount of text Claude can process in a single conversation. Managing context effectively means providing relevant information without exceeding these limits.
Artifacts
Claude's capability to generate visual, interactive outputs directly in conversations. For project managers, this includes Gantt charts, org charts, dashboards, and flowcharts.
Hallucination
When Claude generates specific details—dates, names, numbers—that don't exist in provided context. A common failure mode requiring verification of all outputs.
Drift
The gradual loss of context accuracy over long conversations. Claude may "forget" established constraints or produce outputs that conflict with earlier information.
Framework Terms
CLEAR Framework
A prompt structure for effective AI requests:
- Context: Background situation
- Length/Format: Output specifications
- Examples: Models of good output
- Audience: Who will use this
- Refinement: What to include/exclude
Pre-Mortem
A simulation technique where you assume a project has failed and work backward to identify likely causes. Helps anticipate risks before they materialize.
Devil's Advocate Mode
Instructing Claude to argue against your position or decision. Useful for stress-testing thinking before committing to approaches.
Translation Slider
Adjusting output complexity for different audiences (technical, executive, public) while maintaining core accuracy.
Technical Terms
CLI (Command Line Interface)
Text-based interface for interacting with Claude through terminal commands. Offers speed, privacy, and scripting capabilities beyond the web interface.
MCP (Model Context Protocol)
Protocol enabling Claude to directly access external systems like Google Drive, databases, and APIs. Eliminates manual file transfers.
API (Application Programming Interface)
Programmatic access to Claude's capabilities. Enables building custom applications and automated workflows.
Anthropic
The company that created Claude. Their API provides access to Claude's capabilities for custom integrations.
Token
The unit of text that language models process. Roughly 3-4 characters per token in English. Token limits affect how much context Claude can handle.
Project Management Terms (AI-Enhanced)
Status Report Automation
Using AI to generate formatted status reports from raw notes. Typically reduces report creation from 90 minutes to 10 minutes (review only).
Meeting Processing
AI-assisted extraction of decisions, action items, and summaries from raw meeting notes. Includes follow-up email generation.
Stakeholder Simulation
Using Claude to role-play stakeholder reactions before real conversations. Helps anticipate objections and refine messaging.
Risk Scenario Simulation
Exploring hypothetical risk materializations through AI conversation. Identifies response options and mitigation strategies.
Context Readme
The foundational document in a Project Brain that orients Claude to project basics: objective, status, constraints, current focus, and recent decisions.
Stakeholder Register
A Project Brain component documenting key stakeholders: their roles, concerns, communication preferences, and influence levels.
Decision Log
Historical record of project decisions including who decided, when, rationale, and alternatives considered. Prevents revisiting closed decisions.
Risk Register
Structured tracking of project risks including probability, impact, mitigation strategies, owners, and status.
Workflow Terms
Weekly Capture
The practice of accumulating quick notes throughout the week rather than reconstructing events on Friday. Enables efficient automated reporting.
Prompt Pattern
A reusable structure for AI requests that consistently produces good outputs. Building a library of prompt patterns increases efficiency.
Iteration Chain
The process of progressively refining AI outputs through sequential prompts. Typically involves structure → sections → polish.
Reset
Starting a fresh conversation when context drift has made the current conversation unproductive. Often faster than trying to correct a confused exchange.
Batch Processing
Running multiple similar AI tasks automatically through scripts. Example: generating summaries for all documents in a folder.
Command Center Components
Dashboard
The central view of your command center showing project status, quick actions, and recent outputs.
Quick Actions
One-click buttons that invoke common AI tasks with pre-configured prompts and context.
Output History
Record of generated documents, reports, and communications. Enables review, comparison, and pattern analysis.
Context Manager
Interface for maintaining Project Brain components. Keeps context current without manual file editing.
Quality Assurance Terms
Verification Checkpoint
The practice of checking AI outputs against source documents and known constraints before use.
Red Flag Detection
Recognizing signs of problematic AI outputs: excessive confidence, suspiciously specific details, format mismatch, contradictions.
Trust but Verify
The appropriate mindset for AI outputs: use them as excellent first drafts while maintaining human review for accuracy and appropriateness.
Privacy Terms
Localhost
Running applications on your own machine rather than remote servers. Maximum privacy since data never leaves your computer.
API Key Security
Practices for protecting API credentials: environment variables, never committing to version control, regular rotation.
Data Handling
Understanding what information is sent to AI services and how it's processed. Local installation reduces data exposure.
Skill Development Terms
Prompt Engineering
The practice of crafting effective AI requests. A learnable skill that dramatically improves output quality.
Troubleshooting
Diagnosing and recovering from AI failures. Includes recognizing problem types and applying appropriate recovery techniques.
Workflow Integration
Building AI capabilities into your daily professional routine rather than treating AI as a separate tool.
Using This Glossary
For Review
Scan periodically to reinforce concepts and identify areas for deeper learning.
For Reference
Bookmark for quick lookups when terminology is unclear.
For Teaching
Use definitions when explaining AI project management concepts to colleagues.
For Communication
Ensure consistent terminology when discussing AI capabilities with team members and stakeholders.
Ready to Transform Your Project Management Practice?
This glossary accompanies a comprehensive guide to AI-powered project management. Learn how to save 10-15 hours per week, automate repetitive workflows, and build your own private AI command center.