About

Imported from Onlyaguest/ViviStableSkills (prompt-tuner/SKILL.md). Install upstream with npx skills add Onlyaguest/ViviStableSkills --skill prompt-tuner. Copyright stays with the author.

prompt-tuner

Name: prompt-tuner
Author: Onlyaguest

Treat prompts like code: prompt.txt is the source, cases.yaml is the golden dataset, tuner.py is the test runner, and CHANGELOG.md tracks prompt logic changes.

Use this skill when

The user says "帮我优化这个 prompt"
The user wants prompt regression testing
The project already uses cases.yaml, prompt.txt, or a golden dataset
A prompt behaves badly and needs repeatable debugging
The team wants TDD for prompts instead of one-off prompt edits

Workflow

Run local structure check first.
If files are valid, run real regression tests with configured provider and API key.
Add failing cases before editing prompt rules.
Update version markers in prompt.txt and CHANGELOG.md after changes.

Commands

Local check without API calls:

python3 tuner.py --check

Run full regression:

python3 tuner.py

Optional provider override:

python3 tuner.py --provider openai --model gpt-4o

Inputs and outputs

Input: prompt.txt, cases.yaml, .env
Output: pass/fail table in terminal, plus raw mismatches for failed cases

Validation

Minimum validation command:

python3 tuner.py --check

This checks:

required files exist
prompt.txt has a version header
cases.yaml loads correctly
each case has input and expected

Constraints

Do not change prompt behavior without adding or updating a regression case
Prefer --check for audits and review when API credentials are not configured
Keep prompt version in prompt.txt aligned with CHANGELOG.md

Use it

Copy one of these into your project. Installing also returns the manifest and these snippets.

yaml

targets:
  - https://api.opensmartroute.ai/api/v1/registry/onlyaguest-vivistableskills-prompt-tuner/manifest   # or paste the manifest below

Manifest

An Open Capability Manifest: the router reads it to know what this does, what it costs and when to pick it.

onlyaguest-vivistableskills-prompt-tuner.ocm.jsonjson

{
  "ocm": "1",
  "id": "onlyaguest-vivistableskills-prompt-tuner",
  "kind": "skill",
  "name": "prompt-tuner",
  "description": "Test, tune, and regression-check LLM system prompts with a golden dataset. Use when the user wants to optimize a prompt, run prompt regression tests, manage `cases.yaml`, compare expected vs actual outputs, or follow a TDD workflow for prompts.",
  "publisher": "Onlyaguest",
  "version": "1.0.0",
  "capabilities": {
    "domains": [
      "coding",
      "data_analysis"
    ],
    "tags": [
      "skill-md",
      "prompt",
      "llm",
      "regression",
      "tuning",
      "tdd",
      "github"
    ],
    "languages": [
      "en"
    ]
  },
  "quality_prior": 0.6,
  "examples": [
    "Test, tune, and regression-check LLM system prompts with a golden dataset. Use when the user wants to optimize a prompt, run prompt regression tests, manage `cases.yaml`, compare expected vs actual outputs, or follow a TDD workflow for prompts."
  ],
  "primary": false,
  "metadata": {
    "source": {
      "provider": "github",
      "repository": "https://github.com/Onlyaguest/ViviStableSkills",
      "path": "prompt-tuner/SKILL.md",
      "ref": "27a1421e1a7fa600eb7c2e12e25f1cfe25b7457e",
      "url": "https://github.com/Onlyaguest/ViviStableSkills/blob/27a1421e1a7fa600eb7c2e12e25f1cfe25b7457e/prompt-tuner/SKILL.md",
      "key": "Onlyaguest/ViviStableSkills/prompt-tuner/SKILL.md"
    }
  },
  "instructions": "# prompt-tuner\n\nTreat prompts like code: `prompt.txt` is the source, `cases.yaml` is the golden dataset, `tuner.py` is the test runner, and `CHANGELOG.md` tracks prompt logic changes.\n\n## Use this skill when\n\n- The user says \"帮我优化这个 prompt\"\n- The user wants prompt regression testing\n- The project already uses `cases.yaml`, `prompt.txt`, or a golden dataset\n- A prompt behaves badly and needs repeatable debugging\n- The team wants TDD for prompts instead of one-off prompt edits\n\n## Workflow\n\n1. Run local structure check first.\n2. If files are valid, run real regression tests with configured provi",
  "cost": {
    "context_tokens": 395
  }
}

Fetch it by URL: GET /api/v1/registry/onlyaguest-vivistableskills-prompt-tuner/manifest?version=1.0.0

Reviews

Star ratings from people who tried it. One review per account; edit yours any time.

No reviews yet. Install it, try it, and be the first to rate it.