SDLC.ai VIRIM Agent Navigator

Find the right agent for your VIRIM project phase, role, or deliverable

sdlc-artifact-generator

Skill detail with category, linked agents, and source metadata.

sdlc-artifact-generator

Standardizes artifact output generation for all SDLC agents. Handles JSON-first canonical output with Markdown companion, generated_by watermark injection, VIRIM branding for HTML outputs, and file I/O to the correct docs/ subfolder. Agents delegate output formatting to this skill instead of reimplementing it.

Category: intelligence Used by 19 agents

Source: .github/skills/intelligence/sdlc-artifact-generator/SKILL.md

Used By Agents

Preview

View source preview (first 3000 chars)

# SDLC Artifact Generator

## Overview

Provides a standardized output generation pattern for all SDLC agents. Instead of each agent independently implementing JSON output formatting, Markdown companion rendering, watermark injection, and file I/O, agents invoke this skill's patterns.

## When to Use This Skill

- An agent is producing a deliverable output (JSON report, Markdown companion, HTML dashboard)
- An agent needs to write output files to the correct `docs/` subfolder
- An agent needs to inject a `generated_by` watermark
- An agent is generating a Markdown companion from a JSON canonical output

## Unitary Function

Given an agent name, version, output folder, and structured data, produce:
1. A canonical JSON file with `generated_by` watermark
2. A Markdown companion file with VIRIM branding conventions
3. Files written to the correct `docs/{output-folder}/` path

## Artifact Output Standard

All SDLC agent deliverables MUST follow this standard:

### 1. JSON Canonical Output (REQUIRED)

Every deliverable MUST have a JSON representation as the canonical artifact. The JSON MUST include:

```json
{
  "generated_by": {
    "agent": "<Agent Name from frontmatter>",
    "version": "<Agent Version from body>"
  },
  "report_type": "<AGENT_SPECIFIC_REPORT_TYPE>",
  "generated_at": "<ISO-8601 timestamp>",
  ...agent-specific content...
}
```

**Rules:**
- `generated_by` block is ALWAYS the first key in the JSON root
- `report_type` uses SCREAMING_SNAKE_CASE (e.g., `CODE_REVIEW_REPORT`, `SECURITY_ANALYSIS_FINDINGS`)
- `generated_at` uses ISO-8601 format with timezone
- All other keys follow the agent's specific output schema

### 2. Markdown Companion (REQUIRED for narrative content)

Any artifact that contains substantial narrative text (analysis, recommendations, summaries) MUST also have a Markdown companion. The Markdown MUST:

- Start with a level-1 heading matching the report title
- Include a metadata block at the top:
  ```
  **Generated by:** <Agent Name> v<Version>
  **Date:** <ISO-8601>
  **Report Type:** <Human-readable report type>
  ```
- Use VIRIM font stack for any HTML rendering: "Open Sans", Segoe UI, Arial, Helvetica, sans-serif
- Use VIRIM brand colors for any HTML styling: #00664F (green), #FED141 (gold), #101820 (dark)
- **These values are Context A (document/report HTML outputs) as defined in `AGENT_DEVELOPMENT_STANDARDS.md` Section 8.5. They apply to agent-generated reports in `docs/`. They do NOT apply to React application UI code — React UI uses Context B (`virimTheme`, Roboto, `#FED141` primary). Never mix the two contexts.**
- Include section-level anchors that the JSON can reference
- No emojis in any content

### 3. File Output Convention

All output files MUST be written to the agent's designated output folder:

| Agent Category | Output Path |
|---|---|
| Implementation Planning | `docs/implementation-plan/` |
| BRD Discovery | `docs/brd-discovery/` |
| Change Impact Analysis | `docs/change-impact/` |
| Estima