<?xml version="1.0" encoding="UTF-8"?>
<feed xmlns="http://www.w3.org/2005/Atom">
  <title>PragmAIca Blog</title>
  <link href="https://pragmaica.com/en/blog" rel="alternate"/>
  <link href="https://pragmaica.com/blog/feed.xml" rel="self"/>
  <id>https://pragmaica.com/blog/feed.xml</id>
  <updated>2026-03-03T00:00:00.000Z</updated>
  <subtitle>Articles and updates from PragmAIca: practical AI workflows for development and day-to-day work.</subtitle>

  <entry>
    <title>AI-assisted dev workflow (spec-first)</title>
    <link href="https://pragmaica.com/en/blog/article/ai-assisted-dev-workflow-spec-first" rel="alternate"/>
    <id>https://pragmaica.com/en/blog/article/ai-assisted-dev-workflow-spec-first</id>
    <updated>2026-03-03T00:00:00.000Z</updated>
    <summary>How to structure development around specs and AI tools so you ship faster without losing correctness. Pillar: workflow.</summary>
  <author><name>PragmAIca</name></author>
  </entry>
  <entry>
    <title>AI governance</title>
    <link href="https://pragmaica.com/en/blog/article/ai-governance" rel="alternate"/>
    <id>https://pragmaica.com/en/blog/article/ai-governance</id>
    <updated>2026-03-03T00:00:00.000Z</updated>
    <summary>Policies, risk controls, and auditability for AI-assisted development in enterprises. Pillar: governance.</summary>
  <author><name>PragmAIca</name></author>
  </entry>
  <entry>
    <title>Copilot PR review</title>
    <link href="https://pragmaica.com/en/blog/article/copilot-pr-review" rel="alternate"/>
    <id>https://pragmaica.com/en/blog/article/copilot-pr-review</id>
    <updated>2026-03-03T00:00:00.000Z</updated>
    <summary>Using GitHub Copilot for PR review and consistency with your spec. Cluster: workflow.</summary>
  <author><name>PragmAIca</name></author>
  </entry>
  <entry>
    <title>Cursor rules</title>
    <link href="https://pragmaica.com/en/blog/article/cursor-rules" rel="alternate"/>
    <id>https://pragmaica.com/en/blog/article/cursor-rules</id>
    <updated>2026-03-03T00:00:00.000Z</updated>
    <summary>How to use Cursor rules files to steer AI behavior and keep style and constraints consistent. Cluster: workflow.</summary>
  <author><name>PragmAIca</name></author>
  </entry>
  <entry>
    <title>Evaluation harnesses</title>
    <link href="https://pragmaica.com/en/blog/article/evaluation-harnesses" rel="alternate"/>
    <id>https://pragmaica.com/en/blog/article/evaluation-harnesses</id>
    <updated>2026-03-03T00:00:00.000Z</updated>
    <summary>Automated checks to see if AI outputs meet your spec or quality bar. Cluster: workflow and ROI.</summary>
  <author><name>PragmAIca</name></author>
  </entry>
  <entry>
    <title>Human in the loop</title>
    <link href="https://pragmaica.com/en/blog/article/human-in-the-loop" rel="alternate"/>
    <id>https://pragmaica.com/en/blog/article/human-in-the-loop</id>
    <updated>2026-03-03T00:00:00.000Z</updated>
    <summary>Where and how to require human approval in AI-assisted workflows. Cluster: workflow and governance.</summary>
  <author><name>PragmAIca</name></author>
  </entry>
  <entry>
    <title>LLM observability</title>
    <link href="https://pragmaica.com/en/blog/article/llm-observability" rel="alternate"/>
    <id>https://pragmaica.com/en/blog/article/llm-observability</id>
    <updated>2026-03-03T00:00:00.000Z</updated>
    <summary>Logging, tracing, and auditing for LLM calls in your apps. Cluster: governance.</summary>
  <author><name>PragmAIca</name></author>
  </entry>
  <entry>
    <title>Measuring ROI without lying</title>
    <link href="https://pragmaica.com/en/blog/article/measuring-roi-without-lying" rel="alternate"/>
    <id>https://pragmaica.com/en/blog/article/measuring-roi-without-lying</id>
    <updated>2026-03-03T00:00:00.000Z</updated>
    <summary>How to measure the impact of AI-assisted development with evidence, not hype. Pillar: ROI and metrics.</summary>
  <author><name>PragmAIca</name></author>
  </entry>
  <entry>
    <title>Prompt injection defenses</title>
    <link href="https://pragmaica.com/en/blog/article/prompt-injection-defenses" rel="alternate"/>
    <id>https://pragmaica.com/en/blog/article/prompt-injection-defenses</id>
    <updated>2026-03-03T00:00:00.000Z</updated>
    <summary>How to reduce prompt injection and excessive agency in LLM apps. Cluster: governance.</summary>
  <author><name>PragmAIca</name></author>
  </entry>
  <entry>
    <title>ROI metrics</title>
    <link href="https://pragmaica.com/en/blog/article/roi-metrics" rel="alternate"/>
    <id>https://pragmaica.com/en/blog/article/roi-metrics</id>
    <updated>2026-03-03T00:00:00.000Z</updated>
    <summary>Definitions and practical metrics for AI-assisted development impact. Cluster: ROI.</summary>
  <author><name>PragmAIca</name></author>
  </entry>
  <entry>
    <title>Secure prompting</title>
    <link href="https://pragmaica.com/en/blog/article/secure-prompting" rel="alternate"/>
    <id>https://pragmaica.com/en/blog/article/secure-prompting</id>
    <updated>2026-03-03T00:00:00.000Z</updated>
    <summary>How to design prompts and guardrails so AI outputs stay safe and within policy. Cluster: governance.</summary>
  <author><name>PragmAIca</name></author>
  </entry>
  <entry>
    <title>Spec-first workflow</title>
    <link href="https://pragmaica.com/en/blog/article/spec-first-workflow" rel="alternate"/>
    <id>https://pragmaica.com/en/blog/article/spec-first-workflow</id>
    <updated>2026-03-03T00:00:00.000Z</updated>
    <summary>Practical steps to write a spec before coding and use it with AI tools. Cluster: workflow.</summary>
  <author><name>PragmAIca</name></author>
  </entry>
  <entry>
    <title>Welcome to the LearnAIng Blog</title>
    <link href="https://pragmaica.com/en/blog/article/hello-world" rel="alternate"/>
    <id>https://pragmaica.com/en/blog/article/hello-world</id>
    <updated>2026-03-01T00:00:00.000Z</updated>
    <summary>A short introduction to the LearnAIng blog: talks, bootcamps, and practical AI tools for software teams.</summary>
  <author><name>LearnAIng</name></author>
  </entry>
</feed>
