» Tag
testing
96 postsProbelock: A Lockfile for LLM Tool-Calling Capabilities
Probelock measures an LLM's tool-calling capabilities with deterministic tests and halts CI when a model, quantization, or runtime swap causes a regression.
Evidence Graph turns skipped AI-agent requirements into compile errors
Evidence Graph is a ttsc lint plugin that forces AI coding agents to explicitly cite specs, turning skipped requirements into compile-time errors.
Merging Concurrent Agent Patches by Base Commit, Not Arrival Time
A deterministic merge protocol for concurrent AI coding agent patches, using base commit hashes, diff digests, and hunk ownership instead of arrival time.
CommitBrief — AI code reviews, right in your terminal
A provider-agnostic, local-first CLI that reviews your staged changes, a historic range, or a whole GitHub pull request. Zero telemetry, no server. Free and open source.
commitbrief.comHooks that stop AI coding agents from claiming unverified reads
A hook-based kit blocks AI coding agents from editing unread files, enforcing scoped changes, consumer checks, and verified session handoffs.
Building an 8-Bit RPG With Zero Image or Audio Files
Pixel Quest is a browser RPG with no image or audio assets—sprites are painted from string grids and music is synthesized live via WebAudio.
Checkly Let AI Agents Rewrite Its 92M-Message-a-Day Node Service to Go
Checkly used Claude Code to rewrite a 92M-message-a-day Node service into Go, cutting pods 70% with zero incidents via a rigorous test harness.
How AI-Assisted Debugging Made jsdom 43% Faster for React Tests
Sentry engineers used AI-assisted debugging to fix three jsdom bottlenecks, making a React Testing Library suite 43% faster without changing any tests.
Bringing OS-Style Interrupt Handlers to Browser Automation
How a Cue-based interrupt-handler pattern fixes flaky browser automation scripts, with reliability safeguards and a compile-to-plain-code export.
How a shared JSONL file inflated our eval score to 0.94
A retrieval-based few-shot selector shared its index with the eval set, inflating accuracy from 0.79 to a fake 0.94. Here's the contamination check that catches it.
Loop Engineering: Fixing a Guardrail That Fired Wrong
How to fix a misfiring guardrail check in an agent loop without deleting it—why weakening checks silently disables real catches.