bianhaifeng789-hue/openclaw-config/account-adjustment-strategy
Claude Code and agent skill by bianhaifeng789-hue/openclaw-config
- community
Checking that the SKILL.md file is still in the repository.
Recommend account adjustment actions from ROAS targets, actual recovery, delivery state, and stage. Use this whenever the user asks 怎么调账户, whether to scale/cut/hold, how to adjust budgets, when to stop loss, when not to change too often, or wants practical account动作建议 instead of just model analysis.
Install
npx vskill@latest install bianhaifeng789-hue/openclaw-config/account-adjustment-strategySKILL.md
The same SKILL.md was published earlier as bianhaifeng789-hue/openclaw-config/account-adjustment-strategy.
Account Adjustment Strategy
Use this skill to turn performance state into account actions.
What this skill answers
- Should the account scale, cut, hold, or stop-loss?
- How should budget be adjusted?
- Is delivery being choked by controls?
- Is spend quality too weak to keep buying?
- Should the team wait for more data instead of changing too often?
Expected inputs
Prefer these fields when available:
- recovery_model
- optimization_stage
- target_roas
- actual_d0_roas
- actual_d1_roas
- actual_d3_roas
- actual_d7_roas
- delivery_status
- spend_trend: up / flat / down
- volume_signal: weak / stable / strong
Action definitions
- scale
- increase budget carefully because recovery quality supports expansion
- hold
- keep structure stable and avoid unnecessary noise
- trim
- reduce budget or tighten control because quality is slipping
- stop-loss
- cut aggressively because recovery is too weak with little evidence of catch-up
Core decision logic
Choose scale when
- delivery is stable or constrained for artificial reasons
- later-window ROAS shows healthy catch-up
- model and stage support more volume
Choose hold when
- signal is mixed
- data window is still immature
- recent changes were already significant
- changing again would add noise
Choose trim when
- quality is below target
- later windows help somewhat but not enough
- account still has some value but should not expand
Choose stop-loss when
- multiple windows are weak
- no meaningful catch-up appears
- spend continues but recovery model is breaking
Budget guidance
Use simple action bands by default:
- scale: +10% to +20%
- hold: 0%
- trim: -10% to -20%
- stop-loss: -30% to -100%
Prefer smaller changes in testing and noisy conditions. Prefer moderate changes in scaling. Avoid repeated same-day edits unless the signal is very clear.
Output format
Always respond with:
Recommended action
- scale / hold / trim / stop-loss
- suggested budget adjustment
- main decision window
Why
- recovery reason
- delivery reason
- stage reason
Execution note
- change now or wait
- whether frequent edits are risky
- what to monitor next
Recommended execution
When possible, run the bundled script:
node scripts/decide-account-action.js <json-path>
Use the script for deterministic recommendation, then explain in plain language.
Security
- Security scanPassed, 95 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills-production/account-adjustment-strategy/SKILL.md
Popularity
- 7-day trend score
- 5
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/bianhaifeng789-hue/openclaw-config/account-adjustment-strategy)