一键导入
conveyor-whats-next-nudger
Run a bounded Agent Conveyor "what's next?" nudger loop for an existing Codex app manager/worker pair or worker set.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Run a bounded Agent Conveyor "what's next?" nudger loop for an existing Codex app manager/worker pair or worker set.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Draft, preflight, apply, and inspect Agent Conveyor setup bundles for manager-worker operating cells through a conversational Codex setup flow.
Create one visible Codex app manager and one visible Codex app worker for the current project using the globally installed Agent Conveyor CLI.
Create one visible Codex app manager and multiple visible Codex app workers for the current project using Agent Conveyor.
Run the blocking Agent Conveyor Codex app connection smoke gate for a manager/worker pair or worker set before real work starts.
Wake stale Agent Conveyor Codex app manager or worker threads through native app thread tools and record delivery receipts.
Check Agent Conveyor manager/worker or worker-set status from the current project's per-project ledger.
| name | conveyor-whats-next-nudger |
| description | Run a bounded Agent Conveyor "what's next?" nudger loop for an existing Codex app manager/worker pair or worker set. |
Use this skill when the operator wants an existing Agent Conveyor Codex app manager/worker pair or worker set to continue through a bounded number of manager-approved follow-up passes. This skill does not create sessions, publish, schedule, push, open PRs, merge, or archive workers unless the operator explicitly grants that separate authority.
.codex-workers/workerctl.db under the current project unless the
operator explicitly gives another path.CONVEYOR POLL, CONVEYOR RECEIVED, WORK, CONVEYOR SEND, DISPATCH.mkdir -p .codex-workers
LEDGER="$PWD/.codex-workers/workerctl.db"
conveyor tasks --path "$LEDGER" --json
TASK="example-task"
conveyor app-loop-status "$TASK" --path "$LEDGER" --json
conveyor loop-status "$TASK" --path "$LEDGER" --json
Use the manager's inbox or visible app wake path for the existing setup. If the
setup exposes app-wakeup-plan or app-wakeup-dispatch, prefer that prepared
wake path over ad hoc direct prompts:
conveyor app-wakeup-plan "$TASK" --path "$LEDGER" --json
RUN="loop-run-id-from-loop-status"
NEXT_ITERATION="2"
conveyor enqueue-continue-iteration "$TASK" \
--loop-run "$RUN" \
--requested-iteration "$NEXT_ITERATION" \
--message "Manager approved one more pass. Before acting, print visible CONVEYOR sections, summarize the verified prior evidence, propose the next narrow slice, do the work only if in scope, then notify the manager with receipts." \
--path "$LEDGER" \
--json
conveyor dispatch --watch --watch-iterations 1 --interval 2 \
--dispatcher-id dispatch-local \
--path "$LEDGER" \
--json
If the task is not a Ralph loop, use the existing setup's generated
enqueue-nudge-worker or manager-approved assignment command instead. Do not
invent new shipping authority.
When a role is stale and app-wakeup-dispatch emits send_ready=true, use the
conveyor-app-wake-relay skill to send the prepared wake prompt and record the
delivery receipt. Do not send ad hoc direct prompts.
After each worker pass, require all of:
Stop when the bounded pass count is exhausted, evidence is insufficient, the next proposed slice is out of scope, or the manager recommends stopping. Report a compact receipt with task, ledger, passes used, manager decision, evidence checked, blockers, and the exact next action.
End with:
taskledgerpasses_requestedpasses_completedmanager_decisionevidence_checkedblockersnext_action