Built a small runtime guard for agent tool calls — looking for people to break i

<!-- SC_OFF --><div class="md"><p>I've been thinking about this as local models get better at </p> <p>tool use — shell access, file system, HTTP requests, MCP. </p> <p>If a prompt

<!-- SC_OFF --><div class="md"><p>One recurring issue with domain-specific agents is that overly defensive refusal behavior can make them much less useful once the workflow gets

<ol> <li># Armorer Guard: runtime control should start at the tool call</li> </ol> <p>The more I work on local agent systems, the less I believe static policy alone is enough.</p>

<!-- SC_OFF --><div class="md"><p>I've been following the IETF drafts for agent identity, and I realized there's a massive hole: User-to-Operator trust. We have ways for services

<h2> When Agentic AI Breaks Your Pricing Model </h2> <p><a href="https://github.blog/news-insights/company-news/changes-to-github-copilot-individual-plans/" rel="noopener

<p>Every prompt injection defense I’ve seen makes the same mistake. It asks the wrong question.<br /> The wrong question: “Does this prompt contain dangerous words?”<br /> The

<h2> The Setup </h2> <p>MCP (Model Context Protocol) lets AI agents call external tools. Instead of just reading files and running bash, the agent gets structured access to APIs,

<!-- SC_OFF --><div class="md"><p>Testing agent systems by feeding real natural-language prompts into real runtimes, then scoring whether the correct tool was invoked. No mocks,