메뉴
Forward new information to a running agent, or adjust a task based on changed requirements. Use when the user provides additional context, clarifications, or changed requirements for an in-flight task.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
| name | update-task |
| description | Forward new information to a running agent, or adjust a task based on changed requirements. Use when the user provides additional context, clarifications, or changed requirements for an in-flight task. |
When new information arrives that is relevant to a task that is already running (or waiting), you need to decide what to do with it. The right action depends on how significant the update is.
Figure out which running agent(s) the new information applies to.
If you have multiple agents in flight, the update might apply to one, several, or none of them.
Use mngr list --exclude "has(labels.archived_at)" --exclude "id == \"$AGENT_ID\" --format jsonl and your memory of what each agent is working on to determine this.
If the update doesn't apply to any running task, it might be context for a future task -- store it in memory (if relevant) and move on.
If the new information is supplementary (e.g., a clarification, an additional detail, a pointer to a resource), forward it to the agent via mngr message:
mngr message <agent-id> --message "<additional context>" >& /tmp/`cat /proc/sys/kernel/random/uuid`.message_output &
Note: if the agent is not currently running (e.g., it is stopped or paused), mngr message will start it and deliver the message once it comes online.
This may take a while, so run the command in the background and check that it succeeded afterward (the above command shoves the output somewhere so that it can be inspected later).
If the update fundamentally changes what the task should accomplish (e.g., different goals, different success criteria, "do Y instead of X"), then the running agent's work may no longer be relevant. In this case:
mngr archive -f <agent-id>delegate-task-to-agent skill, incorporating the new information from the startUse your judgment here -- if the agent is nearly done and the change is small enough to course-correct, forwarding a message may be better than restarting from scratch. But if the agent has been working under fundamentally wrong assumptions, a fresh start with correct instructions will be faster and produce better results.
If the user's message makes it clear the task is no longer needed (e.g., "never mind", "don't bother", "I already did it"), archive the agent and move on:
mngr archive -f <agent-id>
Acknowledge the cancellation to the user via send-message-to-user, and check if there is now capacity to launch a pending ticket (via list-tickets).
Submit a bug report to libraries within the `mngr` project (eg, `mngr`, its plugins, `minds`, etc). Use to report bugs you encounter while running that seem to be issues with the `mngr` code itself, rather than with your own code or configuration.
Run commands to explore the current state of mngr and any relevant running agents. Use this skill when you need to debug issues with mngr tasks, or understand what tasks are currently running and their states.
Handle events from the mngr/agent_states source about sub-agent state transitions (finished, waiting, done, etc). You **MUST** use this skill (and *carefully follow the process in this doc*) whenever you receive a message from the "mngr/agent_states" source!
Clean up old agents, their output directories, and other historical data. Use during nightly cleanup or when the system has accumulated stale data.
Handle unexpected situations where the Minds system does not appear to be working as expected. Use when you encounter behavior that seems to contradict what your docs and prompts and skills say should happen, or when something seems broken and you are unsure how to proceed.
Create a sub-agent to perform a task. Use when you need to delegate work to another agent, for example, a working agent (for actually accomplishing some task) or a verifying agent (for deciding what to do about the output of a working agent).