All stories
Essays, launches, case studies, and engineering notes.
246 articlesRSS
All stories, page 2
Page 2 of 10
Launch story: Runtime Governance for AI Agents
Wrap an AI agent's tools in three lines: every call is decided outside the agent against a mandate it can't edit, and a refused call never runs.
Affiliate programs for security tools: what each one actually pays
We run the HOL Guard affiliate program, so we spend a lot of time reading other programs' terms pages. This post is the list we would hand to a writer who covers security tools and wants to know which programs are open, what they pay, and where the gaps are.
Best Grok Plugins in 2026
The best Grok plugins in 2026: 10 picks for Grok Build, sorted by what each one turns on when you trust it, with HOL scan results, verified install commands, example prompts, and pros and cons.

CodeEraser: the write that would bloat your repo is refused before it lands
A Claude Code plugin that refuses a write duplicating code already in the tree, or pushing a file past its line budget, before the file reaches disk.
Launch story: OpenCode Hush
A maintainer's guide to opencode-hush — what it does, who it helps, and how to get started.
How to secure MCP OAuth discovery against SSRF
A practical review plan for hosted MCP clients: validate discovery destinations, bind DNS checks to connections, control redirects and test with mock services.
What is MCP token passthrough? Risks and safer designs
Learn why accepting and forwarding tokens not issued for an MCP server breaks authorization boundaries, with an example, review worksheet and proposed tests.
What are MCP tool annotations? A practical security guide
Understand readOnlyHint, destructiveHint, idempotentHint and openWorldHint, then review MCP tools with concrete examples and a copyable security worksheet.
Best Kimi Plugins in 2026
The best Kimi plugins in 2026: 10 picks for Kimi Code and the Kimi app, with what each manifest loads, what you need before it works, HOL scan findings, verified install commands, example prompts, and pros and cons.
QuietOS: the governor for the hands that hold the keys
QuietOS is a runtime governor for AI agents, with tiered action decisions and a phone-based approval bridge.
Best practices for reviewing AI agent tool calls
An operator checklist for AI agent tool calls: inspect arguments, accounts, data, approval scope and the actual result before trusting an action.
How to test AI agent permissions without production data
Test AI agent permission boundaries with synthetic files, mock services and a six-case worksheet. Separate model refusals from enforced restrictions.
Keep exploring
The engineering desk continues across the ecosystem: security research on shipping agents safely, and plugins that put the standards to work.