| #!/bin/bash | |
| # SPDX-License-Identifier: GPL-2.0 | |
| # | |
| # Reproduce writer stalls after the cgroup that owned the files is removed. | |
| # | |
| # $CG/old appends to NFILES files for OLD_SECS and exits. With MODE=dead | |
| # its cgroup is then removed while the files are still dirty; MODE=alive | |
| # keeps it as a control. $CG/new then appends to the same files for | |
| # NEW_SECS and prints, once a second, its throughput, write latency and | |
| # how far behind schedule it is (lag_s). If bpftrace is available it also |
This isn't a prompting trick. It's infrastructure and habit, built up over months:
Supporting data for golang/go#80899.
This is not a benchmark of CL 692295. It's a minimal, standalone
micro-benchmark isolating just the loop shape question — paired scalar
LDP/STP (copyGP, matching runtime.memmove's current arm64
implementation) vs. a single 4-register NEON VLD1.P/VST1.P per 64-byte
chunk (copyNEON) — with no alignment fixup, no software pipelining, and
no overlap handling. src/dst are fixed, page-resident arenas shared by
| name | clarity-review |
|---|---|
| description | Lint the current diff for the AI-authored-code patterns that repeatedly draw pushback in this repo's PR reviews — comment slop, unverified technical claims, dead code, tautological tests, missing feature flags, single-letter names, unneeded complexity, concurrency footguns, and inconsistent formatting. Run before opening a PR, not after — this is a pre-flight lint, not a bug-catching pass (use hny-review-general/hny-review-thorough for that). |
| disable-model-invocation | true |
Calibrated against roughly 200 real review comments on hound PRs over the prior three months — courtesy of our resident AI slop proliferation curmudgeon — plus recurring #eng-ai/#misc-ai discussion of where AI-authored code and prose fall short.
| Homebrew build logs for lizf/local-mlx/mlx@0.31.1 on macOS 26.4 | |
| Build date: 2026-04-23 14:57:23 |
| Homebrew build logs for lizf/local-mlx/mlx-c@0.6.0 on macOS 26.4 | |
| Build date: 2026-04-23 15:14:12 |
This project modernized the entire firmware stack for the SolidRun HoneyComb LX2K (LX2160A CEx7) from a stale 2021-era codebase to current upstream releases, bringing the existing OP-TEE/StandaloneMm Secure Boot support forward from OP-TEE 3.14.0 to 4.9.0. It is kept current with each EDK2 stable tag; the latest rebase is onto edk2-stable202608 (September 2026).
Result: A fully modern, Secure Boot-enabled firmware booting Ubuntu (24.04 and 26.04 tested) via shim → GRUB → signed kernel.
| Component | Before | After |
|---|
An automated system that generates a daily tech digest from your Bluesky timeline using Claude Code, running every weekday morning at 5am to start your Claude API quota window.
This system fetches your Bluesky timeline, uses Claude to identify and categorize tech-related posts, and generates a beautiful HTML digest that automatically opens in your browser.
This is a comprehensive security audit of the Aurora-Prism ATProto AppView codebase with focus on:
Overall Assessment: The codebase shows strong foundational security with well-implemented authentication, SSRF protection, and parameterized queries. However, there are critical validation gaps that allow malformed data to be persisted, and the Python hook system requires scrutiny as it executes on every user interaction.
| name | bluesky-mcp |
|---|---|
| description | Parse and analyse Bluesky timeline posts and notifications using the Bluesky MCP server. Use when users want to summarise their Bluesky timeline, find technology-relevant posts, extract specific topics from timeline data, or analyse reply/quote notifications. Handles large timeline dumps efficiently and filters for work-relevant content. |
Efficient workflows for analysing Bluesky timeline data and notifications.