Claude Code Stop hook fires per completed turn; SIGKILL mid-generation fires NEITHER Stop nor SessionEnd — bind per-turn side effects to Stop, not SessionEnd

claude-opus-4-8 streams EMPTY thinking deltas via Claude Code — "missing thoughts" is the model, not your display pipeline

Web server launched inside a Claude Code agent Bash tool deadlocks the agent turn (and any orchestrating loop)

Orchestrators spawning CLI agent subprocesses need a sidecar identity map — child transcripts can't be attributed to their owning conversation from content alone

Bun-compiled CLIs (e.g. Claude Code) silently ignore NODE_OPTIONS dns flags — IPv4-forcing must happen at the socket layer

Claude CLI print-mode wrapper stalls or loses history when user settings and session-id reuse conflict

Headless CLI stream-json AskUserQuestion tool-use can be misclassified as an empty response

Edit calls silently lost when concurrent agents share a git worktree on different branches

Thinking-capable local Ollama agents produced empty scorable output in a [redacted:name] benchmark harness

[redacted:name] Qwen3 benchmark agents emit thinking-only output and schema-mismatched findings

In [redacted:name] benchmark runners, auth='none' only removes MCP/graph access u...

[redacted:name] benchmark runner launched duplicate models because per-wave concurrency repeated the wave model

Add local Ollama-backed model trial to a TypeScript benchmark while preserving CLI agent tooling

Add opt-in real [redacted:name] blind crawl e2e for registration discovery

[redacted:name] plugin marketplace tests failed after removing legacy manifest asset arrays

[redacted:name] plugin.json manifest validation fails on CC ≥2.1.123 when agents/commands arrays present

MCP deferred tool index not refreshed after mid-session reconnect — new tools require fresh session

Using MCP Resources to distribute [redacted:name] hooks without git