A developer built a remote Model Context Protocol server for PropHours, a rental-property hours-tracking app, as a single Laravel controller without an MCP SDK. The server implements the streamable HT…
A developer released php2shell-migration, an open-source AI agent skill that helps developers detect and remediate the PHP2Shell vulnerabilities (CVE-2026-109618 and CVE-2026-109619), which allow unau…
A developer released dotvault, a CLI that syncs .env secrets to laptops, CI pipelines, servers and cloud development environments such as GitHub Codespaces, Codex cloud and Claude Code on the web thro…
Nine Point Labs released folderblog 0.2.9, a fix release driven by a full QA pass over the previous 0.2.8 version that followed the published setup instructions as a new user would. The update stabili…
Reelary, a local footage-indexing app, now exposes its search to AI agents over MCP, letting Claude Code assemble a rough cut from a script by calling a rough-cut tool that returns a best clip and two…
A developer field-testing agentwatch, an open-source agent-security tool in the agentic-suite, found that three failing tests were caused by flawed test scenarios rather than broken detectors. In one …
A developer demonstrated that Claude Code's PreToolUse command hooks fail open by default: a hook that times out or exits with an unexpected code does not block the tool call, so a guard meant to stop…
A developer recommends starting with Claude for an EU customer-support chatbot that must return schema-valid structured answers from a private knowledge base, accepting higher latency in exchange for …
A developer and two collaborators built AgentWebUI, a lightweight, self-hosted, agent-first web UI that unifies multiple AI models and command-line agents in a single sidebar. The server runs roughly …
Denis Zontak built a two-service AI pipeline, Night Watch and Day Shift, that detects production errors from logs, triages them with an AI agent, and produces a reviewed pull request with a fix and te…
Savindu Thathsara, a developer in London, reports that his offline AI operating system now correctly parses the number of inputs requested in a natural-language prompt, generating a six-input HTML for…
Langfuse provides an open-source observability and evaluation stack for LLM applications that combines trace ingestion, prompt versioning, dataset snapshots, and eval workflows in a single self-hostab…
Commerce.com deployed WebMCP endpoints across nearly 100,000 storefronts, giving AI agents a standardized tool-discovery interface for browsing, cart management and checkout while keeping per-merchant…
A non-programmer parent used AI to build a site of more than 70 free browser games in four languages, each a self-contained HTML/CSS/plain-JavaScript page with no framework or build step, hosted on Wo…
A developer built "Should I Bother? (The Lazy Bookie)," an open-source MIT-licensed app that turns weather-dependent chores into bets on doing nothing, with a locally hosted Gemma 4 E4B model in Ollam…
A developer has outlined an approach to AI workflow telemetry that cuts operational rework by instrumenting every handoff, retry, latency event and verified business outcome. The method records run ID…
A developer walkthrough shows how ElevenLabs' text-to-speech API exposes emotion tags, prosody controls, and partial SSML support to steer pitch, rate, and volume in generated speech. The example call…
ExchangeRebel launched a pay-per-call API that serves SEC EDGAR Form 4 insider-trading data to AI agents, charging $0.005 per query and $0.01 per changes-since call via the x402 payment protocol on So…
A pre-registered census of PyPI's top 3,000 packages by download count against a seeded random sample of 3,000 from the full 922,283-project index found that 97.4% of head projects declaring an Intend…
Developer Tony describes a verification workflow for code produced by AI coding agents such as Claude Code, arguing that an agent reporting "implementation complete" or "all tests passed" does not est…