The AI Coding Dictionary: Essential Terms for Agentic Software Development

LinkResource

The AI Coding Dictionary from aihero.dev provides plain-English definitions of key concepts in agentic AI coding across seven core domains, spanning models, tooling, failure modes, and workflows.

Key points

  1. Seven sections cover models, sessions and context windows, tools and environments, failure modes, handoffs, memory and steering, and working patterns.
  2. Strict distinction between models and agents: models are stateless next-token predictors, whereas agents combine a model with a harness, tools, system prompts, and context windows.
  3. Handoffs transfer state across bounded sessions using tickets, specs, or lossy context compaction to manage token budgets.
  4. Documented failure modes include attention degradation, hallucination types, knowledge cutoffs, and sycophantic behavior.
  5. Operational concepts range from human-in-the-loop and AFK execution to dark factories, vibe coding, and agent experience (AX).

Why it matters

As software engineering shifts toward autonomous and semi-autonomous coding agents, developers require precise mental models and terminology. Understanding constraints like attention budgets and harness architectures helps teams design reliable agentic workflows.

Analysis

The dictionary published by aihero.dev systematically breaks down agentic mechanics. It avoids hype by focusing on structural constraints: distinguishing parametric weights from loaded contextual knowledge, detailing the trade-offs of compaction, and formalising handoffs between sessions. The taxonomy serves as an architectural glossary for modern AI-assisted engineering practices.

Mentioned links

Found in

Summary generated by AI (Gemini 3.8 Flash, 8 October 2026). It may contain errors – the original source is authoritative.

Related entries

Similar in content, found by the AI search.

  • X post:Matt Pocock

    The AI Coding Dictionary: Clarifying Terminology for AI-Assisted Development

    In response to a call for shared definitions in software engineering, Matt Pocock shared 'The AI Coding Dictionary' (aihero.dev). The resource offers plain-English explanations of key AI coding concepts. In the replies, Pocock noted that he does not distinguish between a specification ('spec') and a plan.

    260bookmarks51.2kviews

    AI & LLMs· Resource

  • X post:Matt Pocock

    Matt Pocock adds Software Factory to the AI Coding Dictionary

    Matt Pocock has added 'Software Factory' to the AI Coding Dictionary. Calling it the hype term of the day, he argues for using the term more accurately and discusses its boundaries alongside concepts like vibe coding and backpressure.

    1.1kbookmarks114.7kviews

    AI & LLMs· Resource

  • Link:IndyDevDan

    Tactical Agentic Coding: Moving Beyond In-the-Loop AI Assistance

    Tactical Agentic Coding (TAC), created by IndyDevDan, is a training framework focused on advancing from basic interactive AI coding to fully autonomous agentic engineering. Rather than remaining 'in the loop' via conversational prompting, software engineers are trained to combine deterministic code with non-deterministic agents like Claude Code to orchestrate self-correcting development pipelines.

    AI & LLMs· Resource

Back to the knowledge base

Let's talk about your project

Get in touch

Locations

  • Mattersburg
    Johann Nepomuk Bergerstraße 7/2/14
    7210 Mattersburg, Austria
  • Vienna
    Ungargasse 64-66/3/404
    1030 Wien, Austria

Parts of this content were created with the assistance of AI.