YAML Frontmatter Reference

Complete reference for YAML frontmatter in prompt and sweep Markdown files.

Prompt File Frontmatter

---
title: "Security Review"
slug: "security-review"
source: "library"
category: "Code Review"
difficulty: "Intermediate"
outputType: "Text"
supportedTools:
  - Codex
  - Claude
  - Cursor
tags:
  - security
  - review
  - owasp
variables:
  - targetDir
  - severity
agentTargets:
  - codex
  - cursor
recommendedPath: ".prompts-gpt/security-review.md"
---

Fields

Field	Type	Required	Description
`title`	`string`	Yes	Display name
`slug`	`string`	Yes	URL-safe identifier (must be unique)
`source`	`string`	No	`"library"` or `"generated"`
`category`	`string`	No	Category for organization
`difficulty`	`string`	No	`"Beginner"` / `"Intermediate"` / `"Advanced"`
`outputType`	`string`	No	Expected output format
`supportedTools`	`string[]`	No	Compatible agent tools
`tags`	`string[]`	No	Search/filter tags
`variables`	`string[]`	No	Declared variable names
`agentTargets`	`string[]`	No	Target agents for sync
`recommendedPath`	`string`	No	Suggested file path

Sweep File Frontmatter

---
title: "Security Audit Sweep"
slug: "security-audit-sweep"
category: "Security"
sweep:
  defaultIterations: 5
  maxIterations: 10
  summaryLines: 40
  strategy: "sequential"
  phase: "security-hardening"
  eval:
    criteria:
      - correctness
      - completeness
      - thoroughness
    passThreshold: 0.8
---

Sweep-Specific Fields

Field	Type	Default	Description
`sweep.defaultIterations`	`number`	3	Default iteration count
`sweep.maxIterations`	`number`	10	Maximum allowed iterations
`sweep.summaryLines`	`number`	40	Lines per summary extraction
`sweep.strategy`	`string`	`"sequential"`	`"sequential"` or `"parallel"`
`sweep.phase`	`string`	—	Phase label for context
`sweep.eval.criteria`	`string[]`	—	Self-evaluation scoring dimensions
`sweep.eval.passThreshold`	`number`	0.7	Pass/fail score threshold (0.0–1.0)

Orchestration Pipeline File

{
  "mode": "pipeline",
  "defaults": {
    "agent": "codex",
    "model": "gpt-5.5",
    "timeoutSeconds": 600
  },
  "steps": [
    {
      "name": "research",
      "promptFile": ".prompts-gpt/orchestrations/pipeline-steps/research.md",
      "agent": "claude",
      "model": "claude-sonnet-4-20250514"
    },
    {
      "name": "implement",
      "promptFile": ".prompts-gpt/orchestrations/pipeline-steps/implement.md"
    },
    {
      "name": "review",
      "promptFile": ".prompts-gpt/orchestrations/pipeline-steps/review.md",
      "agent": "cursor"
    }
  ]
}

Pipeline Step Fields

Field	Type	Required	Description
`name`	`string`	Yes	Step name
`promptFile`	`string`	Yes	Prompt file path
`agent`	`string`	No	Provider override
`model`	`string`	No	Model override
`timeoutSeconds`	`number`	No	Step timeout

Manifest File

manifest.json is auto-generated by sync:

{
  "version": 1,
  "syncedAt": "2026-05-21T14:30:00Z",
  "prompts": [
    {
      "slug": "security-review",
      "title": "Security Review",
      "path": ".prompts-gpt/security-review.md",
      "category": "Code Review",
      "tags": ["security"]
    }
  ]
}

Validation

prompts-gpt validate

Validates all YAML frontmatter in .prompts-gpt/ for syntax and schema compliance.