ConsultingintermediateHITL requiredupdated 2026-08-11

Consulting Assessment

Generate a technical, CRO, AWS, DevOps, or AI-readiness assessment from operating intelligence. Findings grounded in evidence IDs.

#assessment#ai-readiness#aws#consulting

View source →

Tools in the pipeline

api
intel-loader
Load operating objects and evidence for the account.
llm
assessment-model
Emit JSON + Markdown for the chosen assessment type.
human
reviewer
Confirm findings before the client sees them.

Stack: LLM · JSON · Markdown

What it does

Produces a consulting-grade assessment of type technical, cro, aws, devops, or ai_readiness. Separates current-state findings, risks, gaps, recommendations, and roadmap. Grounds findings in evidence IDs when available. From ARC's Operator AI assessment prompt.

Pipeline

Select — assessment type.
Ground — attach evidence IDs to each finding.
Separate — current state, risks, gaps, recommendations, roadmap.
Emit — JSON + Markdown.
Label — mark synthetic accounts as synthetic.

Prompt skeleton

Generate a consulting-grade assessment from the supplied structured operational intelligence.

Assessment types supported: technical, cro, aws, devops, ai_readiness.

Rules:
- Use only supplied facts and explicitly labeled assumptions.
- Ground findings in evidence IDs or operating objects when available.
- Separate current-state findings, risks, gaps, recommendations, and roadmap.
- Never invent confidential client facts.
- Clearly label the account and assessment as synthetic when the source intelligence is synthetic.
- Return: 1) structured JSON  2) polished Markdown.

Guardrails

No invented client facts. Synthetic stays labeled synthetic. Human reviews before delivery. Do not sell a near-term fix as transformation.

Structural dry-run

No model call. Validates required fields only.

Prefills the example payload for AI readiness. Edit it, then run.

← Back to all templates