prompt-to-eval-rubric
Skill0 starsChanged 5 months agoMITView on GitHub
Use when converting prompts, system instructions, assistant behaviors, or policy requirements into eval rubrics with scoring criteria, test cases, assertions, and failure modes.
In this repository19 other files it gives its agents.
- agent-handoff-briefSkill
- api-doc-to-tool-specSkill
- architecture-rfc-reviewSkill
- bug-report-to-reproSkill
- changelog-to-upgrade-planSkill
- competitive-research-briefSkill
- customer-call-to-prdSkill
- database-migration-risk-reviewSkill
- dependency-upgrade-risk-reviewSkill
- incident-log-to-postmortemSkill
- messy-notes-to-decision-logSkill
- privacy-data-flow-mapSkill
- repo-launch-readiness-reviewSkill
- roadmap-to-release-planSkill
- security-review-checklistSkill
- support-ticket-to-evalSkill
- telemetry-to-product-insightsSkill
- tool-permission-auditSkill
- workflow-automation-scoutSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 14
Use when prompt behavior must be evaluated repeatedly or compared across model/prompt versions.
- Distinctive: line 15
Do not use when the task is only rewriting the prompt and no evaluation criteria are needed.
- Distinctive: line 28
Add representative test cases, including edge cases.
- Distinctive: line 50
Test cases include at least one likely failure case.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Prompt To Eval Rubric
Purpose
Turn desired model behavior into a reusable evaluation rubric.
Fit
- Use when prompt behavior must be evaluated repeatedly or compared across model/prompt versions.
- Do not use when the task is only rewriting the prompt and no evaluation criteria are needed.
Inputs
- Prompt, system instruction, policy, desired behavior, or bad output.
- User goals, safety requirements, and known failure modes if available.
- Preferred scoring scale if provided.
Workflow
- Extract expected behaviors and forbidden behaviors.
- Convert each behavior into a measurable criterion.
- Tie each criterion to prompt evidence, policy evidence, or a clearly labeled assumption.
- Add representative test cases, including edge cases.
- Define pass/fail assertions and scoring guidance.
- Preserve uncertainty where behavior is underspecified.
Output
Produce Markdown with:
- Eval Goal
- Rubric Table
- Test Cases
- Assertions
- Failure Modes
- Scoring Notes
- Missing Requirements
Validation
- Criteria are observable in outputs.
- Assertions can be judged consistently.
- Safety and correctness are not traded away for style.
- Ambiguous prompt requirements become missing requirements.
- Test cases include at least one likely failure case.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

