AgentSkillsCN

tuning-proposal-apply

安全地应用确定性调优建议动作（目前：新增/更新v2负面关键词）。

SKILL.md
--- frontmatter
name: tuning-proposal-apply
description: 'Safely apply deterministic tuning proposal actions (currently: add/update
  v2 negative keywords).'
user-invocable: false
allowed-tools:
- Bash
- Read
- Write

tuning-proposal-apply

When to use

  • •You have a reviewed tuning proposal and want to apply safe config edits.

Inputs

  • •proposal YAML path.
  • •apply flag (otherwise dry-run).

Workflow

  1. •Dry-run: python -m ops.cli quality apply-tuning --proposal /tmp/proposal.yaml.
  2. •If output looks correct, apply: python -m ops.cli quality apply-tuning --proposal /tmp/proposal.yaml --apply.
  3. •Commit the config changes and measure impact with quality stats.

Outputs

  • •JSON summary with applied/skipped actions.

Guardrails

  • •Auto-apply scope is intentionally narrow. For broader changes, treat the proposal as guidance and make PRs.

References

  • •references/reference.md
  • •docs/QUALITY_OPS_ARCHITECTURE.md