Fetching from the wire…
01
02
03
04
05
06
07
08
Public story · 2026-09-06 · high
The manifest bumped a version number the build never rebuilt, and a version check turned that mismatch into a kill loop.
Why now: The fix went up as v13.24.1, covered September 6.
claude-mem's 13.24.0 release bumped the manifest version but skipped the rebuild, shipping .cjs files that still held 13.23.1 bytes. That gap put the worker into a crash loop, and every hook event killed the in-flight observer before it could finish. Anyone running claude-mem lost whatever the observer was capturing mid-run, over and over, with no error, just another respawn.
ensureWorkerRunning() checks the plugin-cache directory's version against the worker's baked-in __DEFAULT_PACKAGE_VERSION__. When they don't match, it SIGKILLs the worker and respawns it. The respawned file was the same stale build, so the check failed again on the next hook event.
The v13.24.1 fix ships the rebuilt artifacts, closing the gap between manifest and bytes.
The bug isn't unique to this plugin. Any Claude Code plugin distributed as committed build output, rather than built at install time, has the same failure mode available. Bump the version string without rebuilding, and it can land in the same loop.
The manifest version and the compiled bytes are two separate things a release has to get right. A git commit doesn't check that they match. A CI step that fails on a version bump with no built-file change would catch this before it ships.
Each link below shares sources, entities, or timing with this story.
Rust, created May 14, at 2,643 stars (GitHub). Every run produces checkpoints linking a commit to the session that made it, including prompts, tool calls and reasoning. It runs Claude Code, Codex, its own agent and anything from the ACP registry side by side against one codeba...
A spec is a press release until someone who didn't write it implements it. GitHub made Agent Plugins 1.0 generally available on August 12 across VS Code, Copilot CLI, the Copilot SDK, and the Copilot app on all plans. The spec, published August 6, was co-authored by AWS, Anysp...
wanshuiyin/HERO-Anti-OverDefense went from creation to 68 stars in a single day. HERO is Hashing, Edge cases, Rubrics, Overbuild, and the claim is that agent over-engineering isn't diffuse but falls into four recognizable shapes suppressible with a portable prompt contract acr...
The TypeScript project compiles a repository into an explorable, searchable graph you can ask questions against, with the explicit framing "graphs that teach beat graphs that impress." It targets Claude Code, Codex, Cursor, and Copilot. Directly relevant if your agent currentl...
This is the paper of the week. arXiv 2607.28871 introduces BSG-VA, which replays every validation command an agent runs across three code states: the original buggy code (B), the candidate patch (S), and the gold developer fix (G). If a test passes in all three states, it neve...
This Rust harness (+2,585 stars) competes on resource footprint rather than features: 27.8 MB PSS for a single session with local embedding disabled, claimed 13.9× less than Claude Code and 6× less than jcode's own embedding-enabled mode. Time-to-first-frame 14.0ms against a c...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.