config-workflow

name

config-workflow

description

OpenClaw configuration management safety protocol and change reporting. Use when modifying OpenClaw configuration that affects runtime behavior: (1) editing openclaw.json (bindings, channels, agents), (2) changing workspace prompt files (SOUL.md, AGENTS.md, IDENTITY.md), (3) restarting gateway or agents, (4) installing/removing skills or tools, or (5) any action that may impact message routing or agent behavior.

Configuration Change Workflow

This skill provides safety protocols and structured reporting for OpenClaw configuration changes.

Core Principles

Report all runtime-affecting changes. Any configuration change that impacts message routing, agent behavior, or service availability requires proactive reporting in the group chat.

Never rely on memory. Configuration knowledge evolves. Always verify against source code, documentation, or the shared knowledge base (kb/).

Follow the change sequence. For high-risk operations: git status → execute → report → rollback info.

---

Change Reporting Structure

When you complete a configuration change, bug fix, restart, or any operation that affects runtime state, report in the group chat using this structure:

Four Required Sections

1. 做了什么改动 (What changed)

   • Which files/parameters were modified
   • Before/after summary (if meaningful)

2. 如何验证 (How to verify)

   • Validation commands: jq checks, openclaw doctor, openclaw status
   • Functional test: trigger a test message, verify routing works

3. 当前是否已生效 (Current status)

   • Is restart/hot-reload needed? Already done?
   • Effective scope: which agents/channels affected?

4. 风险 & 下一步 (Risks & next steps)

   • Do you need user confirmation?
   • Rollback point/backup available?

Example Report

• 做了什么：把 ginmoni 的绑定从 allowlist 改为 groups 模式，移除 3 个废弃的 group_id
• 如何验证：openclaw gateway call config.get | jq '.channels.telegram.accounts["8402020404"]'
• 是否已生效：已重启 gateway，所有使用 ginmoni 的群组已应用新规则
• 风险：无（allowlist → groups 是放宽限制）；回滚点：git ref 62f3731

For more examples, see EXAMPLES.md.

---

Safety Protocol

Before Critical Actions

Before executing critical actions (configuration changes, deployments, irreversible operations):

1. Search kb/ for known issues and playbooks:

   kb/ | grep -i "Known.*Issues"
   kb/ | grep -i "Playbook.*<topic>"
   

2. Follow documented protocols if found:

   • Read the playbook/issue document
   • Apply the recommended steps

3. Ask first if uncertain:

   • When no documentation exists
   • When the operation is high-risk
   • When the outcome is uncertain

Verification Sources

Always verify against authoritative sources:

• Source code (for OpenClaw internals)
• Documentation (docs/, skills/*/SKILL.md)
• Knowledge base (kb/ for playbooks and known issues)

---

Change Sequence

For high-risk operations, follow this order:

1. git status — Confirm no accidental deletions/modifications; commit what should be saved
2. Execute — Perform the risky/irreversible operation (delete/move/overwrite)
3. Report — Provide structured report (see above)
4. Rollback info — Always include recovery steps

---

Integration with AGENTS.md

This skill complements the safety guidelines in AGENTS.md:

• AGENTS.md: Universal safety principles (trash > rm, ask before external actions)
• This skill: Configuration-specific workflow (change reporting, kb/ protocols)

When working with OpenClaw configuration:

1. Follow AGENTS.md safety rules
2. Apply this skill's reporting structure
3. Use tools/tg-compiled.sh and tools/tg-routing-map.sh for validation

---

References

• EXAMPLES.md - Detailed reporting examples for common operations
• PLAYBOOKS.md - Common configuration change patterns and troubleshooting