菜单
Act on the results of a verifying agent by reviewing its recommendations and taking the appropriate next actions. You **MUST** use this skill (and *carefully follow the process in this doc*) whenever a verifying agent finishes!
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
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).
| name | handle-verification-result |
| description | Act on the results of a verifying agent by reviewing its recommendations and taking the appropriate next actions. You **MUST** use this skill (and *carefully follow the process in this doc*) whenever a verifying agent finishes! |
When a verifying agent finishes, read its verdict.json and next_steps.md files, and execute the recommended actions.
cat output/<verifier-agent-id>/verdict.json
This file contains the verdict (PASSED / FAILED), confidence, and reasoning, which helps contextualize your next steps.
Next, read the next_steps.md file:
cat output/<verifier-agent-id>/next_steps.md
These are the actions that were recommended by the verifying agent.
Think carefully about the verdict.
If you disagree, you can send a message to the verifying agent (by calling mngr message <verifying-agent-id> --message "Your message goes here") to ask for clarification, include new information, or point out anything you think it missed.
In such a case, you should stop executing this skill, and wait for the verifying agent to update its verdict (which will ultimately trigger this skill again)
Remember that a "FAILED" verdict does not necessarily mean that the work was bad or that the user will be unhappy--it just means that the verifying agent found some issue that it thinks should be followed up on.
Use your judgment to decide whether to challenge the verdict or move on to executing the next steps.
If not, and the verdict seems reasonable to you, then move on to the next step.
Assuming the verdict and next steps seem reasonable, next you should work through the next steps list one by one. These may include:
git merge, shell commands)send-message-to-user)create-ticket)mngr archive)tk close)delegate-task-to-agent)Post each action to the Work Log as you execute it. For example:
Verification of <task-name>: PASSED.
- Merged branch feature/fix-auth
- Closed ticket tk-5c46
- Archived agents agent-abc and agent-def
- Sent completion notice to user
If any important data or artifacts need to be preserved beyond the output directory (e.g., a report the user should keep, configuration that should be committed), move or copy them to a permanent location now--the output directory will be cleaned up when the agents are eventually destroyed.
After executing the next steps, you should also double-check if anything was missed by the verifying agent's recommendations, for example:
mngr archive -f <agent-id>)