Expose stale-running task maintenance details in CLI
TL;DR
The tasks maintenance CLI command now outputs detailed JSON explanations for retained and reconcile candidates, detailing backing-session, cron, CLI, and wedged-subagent states.
What changed
The tasks maintenance CLI command now outputs detailed JSON explanations for retained and reconcile candidates. It surfaces backing-session, cron, CLI, and wedged-subagent states in one structured response.
This update landed in the current OpenClaw release on GitHub. No price change applies since the tool remains free under the MIT licence.
Why it matters
Vibe Builders gain a direct window into why scheduled tasks linger or fail instead of guessing from partial logs. The change pressures closed agent platforms that hide similar state data behind dashboards.
It bets that transparent self-hosted runtimes will win over users who already manage their own VPS and LLM spend. Persistent agents only stay useful when operators can spot and clear stuck processes quickly.
How to use it
Pull the latest OpenClaw code from the GitHub repository and rebuild the CLI binary. Run the command openclaw tasks maintenance to receive the new JSON output.
Pipe the result into jq or a simple script for alerts. The feature works on any supported platform once the update is installed.
Watch for
Adoption will show in community scripts that parse the JSON for automated cleanups. The bet breaks if the exposed states remain incomplete for common failure modes. Expect similar JSON detail to appear in the heartbeat and browser-control commands next.
Who this matters for
- Vibe Builders: Use the new JSON output to debug stuck agents without digging through messy system logs.
- Developers: Pipe the maintenance CLI output into jq to automate cleanup of wedged subagents and stale sessions.
Harsh’s take
OpenClaw is leaning into the transparency that closed-source agent platforms lack. By exposing granular JSON states for wedged subagents and cron tasks, they are handing operators the telemetry needed to run reliable, long-running agents on their own hardware. This is a direct play for the self-hosting crowd who prioritize control over convenience.
The move to structured output suggests a shift toward programmatic management of agent fleets. If you are building agentic workflows, the ability to see exactly why a task is retained or reconciled is the difference between a production-ready system and a fragile prototype. Expect this level of detail to become the baseline for open-source runtimes.
by Harsh Desai
About OpenClaw
View the full OpenClaw page →All OpenClaw updatesMore from OpenClaw
- App UpdateAdd doctor warnings for hidden MCP server tools
The doctor utility now warns users when sandbox tool policies hide configured Model Context Protocol (MCP) server tools before provider requests are made.
- FeatureSupport per-agent lean local-model configuration
Added support for the experimental localModelLean configuration option on individual agents. This allows lean local-model mode to be enabled selectively rather than globally.
- FeatureIntroduce bundled Policy plugin for workspace conformance
Added a bundled Policy plugin that enables policy-backed channel conformance checks, doctor lint findings, and opt-in workspace repair to streamline configuration management.