safe-write-operations
@markifact · 收录于 1 周前
Reference — rules for safely executing write/destructive operations against ad accounts. Always loaded into the performance-marketer agent.
适合你,如果管理广告账户时需要安全执行删除或修改操作
/ 下载安装
用别的 agent?下载 .zip 解压,把文件夹放进它的技能目录
Claude Code
~/.claude/skills/(项目级 .claude/skills/)Codex CLI
~/.codex/skills/Cursor自动读取上面两处目录
其他工具见其文档的「skills」目录;两个下载是同一份文件,只是名字不同
/ 通过 npx 安装 校验哈希
npx oh-my-skill add markifact/markifact-mcp/safe-write-operations/ 通过 bash 安装
curl -fsSL https://oh-my-skill.com/install.sh | bash -s -- markifact/markifact-mcp/safe-write-operations/ 已经装过?验证本机副本,不用重装
npx oh-my-skill verify markifact/markifact-mcp/safe-write-operations安装目标可用 --agent / --scope 或 --to 明确指定;省略时只会在唯一已存在的 agent 目录上自动选择,零命中或多命中会停止并提示。content_hash 缺失或不一致均拒装。
49GitHub stars
~775上下文体积 · 单文件
镜像托管
怎么用
技能原文 SKILL.md
Safe write operations
Calls to run_write_operation can spend money, pause campaigns, change budgets, edit live creatives, send emails, or charge customers. Treat every one as a production change.
When this protocol applies
The trigger is the requires_approval: true flag on the operation in the find_operations response. If requires_approval: true, you must use run_write_operation and must follow the four-step protocol below. If requires_approval: false, use run_operation and skip this protocol — no confirmation needed.
The four-step write protocol
Always follow this exact sequence before calling run_write_operation:
- State the change in plain English. Include the platform, account, object name/ID, and what will change. Example: > "I'm about to pause the Google Ads campaign 'Brand — US Search' (ID
1234567890) in account Acme US (MCC 999-888-7777). This will stop spend immediately." - State the blast radius. What spend, traffic, revenue, or audience is affected? Pull a quick number from
run_operationif you don't know. - Ask for explicit confirmation. Wait for a yes. "Looks good", "go ahead", or "do it" all count. Anything ambiguous = ask again.
- Execute, then verify. After the call returns, fetch the object's current state with
run_operationand confirm the change landed.
Hard rules
- Never batch more than 5 write operations without re-confirming. If a workflow needs 20 pauses, do the first 5, summarise, ask "continue with the next batch?".
- Never delete unless the user said "delete" (not "remove", not "pause"). Prefer pause/disable over delete every time.
- Never change budgets by more than 50% in a single call without an extra confirmation step. Big budget moves are easy to fat-finger.
- Never run a write operation immediately after a model-invoked discovery. The user must have asked for the change in their own words first.
- Never assume a connection. If the user has multiple Google Ads accounts and didn't specify, ask.
Always check before executing
- "Will this reset learning?" On Meta, budget shifts > 20%, audience changes > 10%, placement changes, optimization-event changes, and major creative swaps reset the ad set's learning phase (~7 days of degraded performance). On Google Ads, switching bid strategies resets learning (~1–2 weeks). Warn the user explicitly when a planned change will trigger a reset.
- "Does this change require fresh ID lookup?" Structural changes (creating a new RSA, replacing a Meta creative, creating a new ad set) return new IDs. Don't reuse stale IDs from earlier in the conversation after a structural change — re-fetch.
- "Is there a dedicated op for this?" On Google Ads,
gads_mutateis the last-resort generic. If a dedicated op exists (e.g.gads_update_campaign_budget), use it — clearer intent, better validation, fewer footguns.
When run_write_operation errors
- Auth error → user must reconnect at <https://www.markifact.com/app/connections>. Stop, don't retry.
- Validation error → re-fetch the operation schema with
get_operation_inputs, fix the payload, ask the user to confirm the corrected call before retrying. - Rate limit / quota → back off, do not retry in a loop.
按 MIT 许可原样转载,未经改动 · 在 GitHub 查看 →
评论
登录即可评论;带「已验证安装」的,是发布者名下有本店的安装或持有记录。
…