Long running tasks
Linlin's curated AI agent harness configuration: workflow rules, skills, hooks, plugin recommendations, tooling preferences, and project templates. Loadable into any new project so a fresh /init can pick the relevant subset.
npx -y skills add jajupmochi/agent-harness --skill long-running-tasksAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Manage long-running commands and multi-step work in Codex. Use when a task may exceed normal turn time, needs a dev server, watcher, training run, crawl, benchmark, or repeated polling.
SKILL.md
1.2 KB, as published. Nobody here has run it
long-running-tasks
Read ../general/long-running-tasks/SKILL.md for the original policy, then use
these Codex mappings:
- Claude background Bash -> Codex background
exec_commandsession. - Claude Monitor -> periodic
write_stdinpolling or explicit status command. - Claude subagent -> Codex subagent only if available and useful; otherwise use a focused plan and background terminal.
- Claude "ask before long run" -> Codex approval when the run needs network, writes outside the workspace, high cost, or destructive side effects.
Workflow:
- Estimate duration and side effects.
- For commands expected to run longer than a normal tool call, start them in a background session and keep the session id.
- Poll at reasonable intervals. Do not spam the transcript with full logs.
- Stop or clean up sessions that are no longer needed before final response.
- If a task is too large for one run, leave a concrete continuation state: current command, session id, last observed output, and next check.