From the counter
52 posts
-
KEV first, EPSS second, everything else on schedule: a written prioritisation policy for teams that will never patch 66,000 CVEs
With about 66,000 CVEs projected for 2026 and exploited ones well under 1%, rank by known exploitation, then by EPSS probability, and give each tier a written deadline. Template policy text included.
-
You can't govern what you can't list: an inventory of skills, MCP servers and hooks across developer laptops
Agent add-ons live in per-user JSON files and folders that MDM, EDR and SIEM do not parse. Build the inventory from those files on every laptop and every checkout. Config-path map dated 15 October 2026, plus a collection script.
-
The intel commons is rationed: what abuse.ch, VirusTotal, NVD and MS-ISAC changes mean for teams built on free
Between November 2024 and April 2026, four sources small teams treat as free added mandatory authentication, non-commercial tiers, paid membership or triage. Free intel now costs accounts, terms reviews and fallbacks.
-
Five Claude Code hook settings that quietly turn agent enforcement into advice
Exit 1 does not block. A timed-out hook does not block. The if filter is best-effort. Plugin tools carry prefixed names. User hooks merge in unless managed-only mode is on. Check all five before you call a hook a control.
-
Fake OpenClaw installers ranked by AI search: a postmortem and a five-minute check for new agent tools
In February 2026 Bing's AI search pointed a developer to a fake OpenClaw installer on GitHub. Five checks, about a minute each, would have stopped it. Here is the timeline and the routine.
-
Egress allowlisting for coding agents and CI runners: the control that would have stopped the axios callback
Most agent-ecosystem malware has to phone home. An outbound allowlist on CI runners and agent sandboxes contains compromises that known-bad lists miss. Start in audit mode, then block. The ongoing cost is keeping the list current.
-
Treat the developer laptop as a privileged workstation: what agent tokens and stealers changed
Microsoft's device model puts general developers in the enterprise tier, where users install their own software. Agent credentials, MCP configs and stealers aimed at developers make that the wrong tier. Includes a tier test.
-
What one compromised developer laptop costs in the agent era: a P10/P50/P90 model you can adapt
A model built on public wage data puts the internal response cost of one agent-related laptop compromise at about $2,000 to $46,000. What moves it toward a claim is whether a stolen credential is used before rotation.
-
Do coding agents call optional security tools on their own? What the docs guarantee, and a 50-task test to run
Nothing in the MCP spec or Claude Code's docs makes an agent call a security-check tool before it installs something. The model decides. Here is why, and a pre-registered 50-task protocol to measure it on your own setup.
-
Two-thirds of AI users at work use personal accounts: moving coding agents to managed tenancy without a ban
DBIR 2026, via CRI, says 67% of AI users reach AI through non-corporate accounts on work devices. For coding agents that means no org policy on tools that run commands. Make the managed path easier, then enforce it on the device.
-
Before you approve coding agents org-wide: a seven-condition rollout gate for security leaders
Approve an org-wide coding-agent rollout when seven verifiable conditions hold - inventory, managed settings, a deliberate default mode, an artifact policy, egress limits, a decision log and a tested runbook. Most take days, not months.
-
ClawHavoc was a registry-governance failure: what 341 malicious skills say about skill stores and takedowns
ClawHavoc is retold as a malware story. The useful reading is governance: an open store let a few publishers ship hundreds of look-alike skills. Publisher limits, instruction review and cross-store matching matter more than the payload.
