Who contributed to AI Agents on August 30, 2026?

5 developers shipped this update, including Sourav, ccurme, mkondratek, Logan, and dependabot.

What were the notable AI Agents updates?

fix(vscode): thread task id into hook runner creation so execution telemetry fires, fix(vscode): thread task id into hook runner creation so execution telemetry fires (#13547), and fix(vscode): prevent hook spawn failures from crashing the core process.

Aug 30, 2026

$ the-wire · showcase

Cline fixes hook telemetry and crashes; LlamaIndex purges unfixable packages

By RepoJournal · Filed 07:01 UTC on August 30, 2026 · About AI Agents

5 people shipped this

mkondratek @mkondratek 4 cited

Sourav @Sourav1331 1 cited

Logan @logan-markewich 1 cited

dependabot @dependabot 1 cited

ccurme @ccurme

Cline's VS Code extension now fires hook execution telemetry and survives hook spawn failures, while LlamaIndex removes packages with unfixable vulnerabilities from main.

Cline's next VS Code extension emitted zero `hooks.execution` telemetry events because `apps/vscode/src/sdk/hooks-adapter.ts` created hook runners without a task id, and `StdioHookRunner` gates `captureHookExecution` on `if (taskId)`. [1] The fix threads the task id (and tool name for tool hooks) through all five `factory.create` call sites, pinning it with a regression test. [2] Separately, a hook child-process spawn failure emitted an "error" event with no listener, which Node's EventEmitter turned into an uncaught exception, killing the entire cline-core process with exit code 1. [3] Guarding the emit behind `listenerCount` lets the rejection propagate to `StdioHookRunner` without crashing the core. [4] In llama_index, a sweep of security alerts led to both dependency bumps and removal of packages whose remaining vulnerabilities cannot be fixed from this repo (abandoned upstreams, dead APIs, or unfixable pins); these are preserved on a `stale_packages` branch. [5] The removed packages include `llama-index-callbacks-aim`, `llama-index-callbacks-agentops`, `llama-index-llms-palm`, and the `llama-index-readers-airbyte-*` family. [5] A dependabot PR bumps pytest, pillow, and requests across three integration directories. [6] LangChain fixed a `create_agent` conditional edge routing bug where middleware-injected synthetic `ToolMessage` objects caused LangGraph to raise `KeyError("model")`; the fix includes `loop_entry_node` in `model_to_tools_destinations` with a regression test. [7]

Action items

→ If you use Cline hooks, upgrade to the latest VS Code extension to enable hook execution telemetry and prevent core crashes cline/cline [immediate]
→ If you depend on any of the removed llama_index packages, migrate before your next deploy run-llama/llama_index [immediate]
→ If you use LangChain create_agent with middleware that injects synthetic ToolMessage objects, upgrade langchain to get the edge routing fix langchain-ai/langchain [plan]

References

[1] fix(vscode): thread task id into hook runner creation so execution telemetry fires ↗ cline/cline
[2] fix(vscode): thread task id into hook runner creation so execution telemetry fires (#13547) ↗ cline/cline
[3] fix(vscode): prevent hook spawn failures from crashing the core process ↗ cline/cline
[4] fix(vscode): prevent hook spawn failures from crashing the core process (#13422) ↗ cline/cline
[5] chore: remove stale packages from main (2026-08 dependabot cleanup) ↗ run-llama/llama_index
[6] chore(deps): bump the uv group across 3 directories with 3 updates ↗ run-llama/llama_index
[7] fix(langchain): include model destination in agent tool routing ↗ langchain-ai/langchain