Calibrate
ProductivitycalibrateCalibrate is an Agent Skill for productivity published in the sharpdeveye/maestro repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.
Use when workflow components are inconsistent, naming conventions vary, or a new team member's work needs alignment to project standards.
How to use this skill
Bring this guide into your coding agent with a prompt tailored to the tool you use.
- Open your project in Codex.
- Copy and paste the prompt below.
- Review files and risks before approving any action.
Inspect the Agent Skill at https://github.com/sharpdeveye/maestro/blob/HEAD/source/skills/calibrate/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/calibrate, but wait for my approval before writing files or running scripts.
Copying this prompt does not install or run the skill. Review third-party files before use.
MANDATORY PREPARATION
Invoke /agent-workflow — it contains workflow principles, anti-patterns, and the Context Gathering Protocol. Follow the protocol before proceeding — if no workflow context exists yet, you MUST run /teach-maestro first. Consult the prompt-engineering reference in the agent-workflow skill for naming and style consistency patterns.
Ensure consistency across all workflow components. Inconsistency creates confusion — for the model, for developers, and for users.
Calibration Dimensions
Naming Conventions
- Tool names follow consistent pattern (verb_noun, noun.verb, or camelCase — pick one)
- Agent names follow consistent pattern
- Configuration keys follow consistent pattern
- File names follow consistent pattern
Prompt Style
- All prompts use the same structural pattern (4-zone)
- Consistent delimiter style (XML tags, markdown headers, triple-dash)
- Consistent output schema format (JSON schema, markdown template)
- Consistent instruction style (imperative, numbered steps)
Error Handling
- All tools return errors in the same format
- Error codes follow consistent scheme
- Error messages follow consistent tone
- Retry logic uses consistent strategy
Logging
- All logs use the same format (JSON structured, text, etc.)
- Consistent field names across all log entries
- Consistent log levels (debug, info, warn, error)
- Consistent PII redaction approach
Calibration Process
- Identify the standard: What's the most common pattern in the existing codebase? That's the standard.
- List deviations: Find all components that deviate from the standard.
- Prioritize: Fix the most impactful deviations first (user-facing > internal).
- Apply: Make the changes, ensuring tests still pass.
- Document: Update
.maestro.mdwith the established conventions.
Consistency Audit Table
| Dimension | Standard | Deviations Found | Priority |
|---|---|---|---|
| Tool naming | ? | ? of ? tools | High/Med/Low |
| Prompt structure | ? | ? of ? prompts | High/Med/Low |
| Error format | ? | ? of ? tools | High/Med/Low |
| Log format | ? | ? of ? entries | High/Med/Low |
Calibration Checklist
- Convention standard identified for each dimension
- All deviations listed with location
- Highest impact deviations fixed first
- Tests pass after each calibration change
- Updated
.maestro.mdwith established conventions
Recommended Next Step
After calibration, run /refine for a final polish pass, or /evaluate to verify consistency improvements.
NEVER:
- Invent new conventions when existing ones work
- Calibrate in a way that changes behavior (this is standardization, not refactoring)
- Skip test verification after calibration
- Change naming conventions without updating all references