A security review of an unnamed AI coding agent governance tool found five live defects, including three security-relevant ones, despite 254 passing tests, clean clippy checks, and five green CI jobs.…
A Rust workspace's AI coding agent hook failed to enforce a security rule because the `clippy::let_underscore_must_use` lint is off by default, allowing a session that read untrusted data to reach the…
Socket, a supply-chain security scanner, flagged the npm package ai2rules-harness with a Supply Chain Security score of 64%, prompting its developer to discover that the package's postinstall script f…
A coding-agent security tool shipped by an unnamed developer was found to have three critical flaws, including a kill switch at `.claude/gate-off` that the agent could create without denial, allowing …
Ai2rules-harness, a new open-source tool, lets developers govern coding agents with a single command by installing a gate that denies tool calls based on destination path and taint state, not command …
A new governance scorecard for coding agents, published by an unnamed tool builder, found that its own product fails three of nine control questions, including whether its hook can grant permission in…
Microsoft's Agent Host Protocol, an open-source MIT-licensed specification at version 0.7 with a reference implementation in VS Code, defines how AI agents request human approval for tool calls, but i…
A team that built a benchmark tool for AI-text detectors found that it caught two of their own wrong conclusions within a week, including a claim that a simple detector performed below chance based on…
LangChain published a post arguing that agents need their own computer, and that a sandbox does not stop prompt injection but does contain the execution blast radius. In response, the AI startup (whic…
OpenAI's essay on harness engineering describes a small team that shipped a real product with no hand-written code, as Codex generated the application, tests, CI, documentation, and internal tools, gr…
A live Claude Code session governed by the kernel granted requests to read /etc/shadow and write ~/.bashrc exactly as it granted reading the project's own README, revealing that the governance manifes…
A developer at a governance tool project found that its own image-generation tool had a remote-code-execution vulnerability because it ran an agent with `--dangerously-skip-permissions`, auto-approvin…
Anthropic's Claude Code permission system is vulnerable to prompt injection because its flat allowlist cannot track the provenance of tool calls, a concept known as 'taint' that has existed since Perl…
A security paper by Fan et al. (2026) proves that flat tool-level monitors, which track taint as a single label per session, must either let dangerous actions through or block harmless ones. The team …