HyperAgent Ops Copilot

Five saved agents, all bounded by human approval.

With `$1000` in HyperAgent credits, use a stronger Day 1 setup: weekly research and monitoring plus per-release QA. Agents can draft and test; they cannot deploy, change pricing, publish indexed pages, or send outreach without approval.

Market Intel Agent

Cadence: Weekly

Budget cap: $60/run

50 transformation keywords, DataForSEO metrics, SERP notes, and top 10 candidates.

Spec Agent

Cadence: Per utility page

Budget cap: $35/run

PRD, API behavior, UX copy, acceptance criteria, and issue breakdown.

QA Agent

Cadence: Before deploy

Budget cap: $45/run

Desktop/mobile screenshots, tool-state checks, broken links, and release recommendation.

Launch Agent

Cadence: Per launch

Budget cap: $40/run

GitHub README, HN post, Reddit posts, X thread, directory copy, and marketplace listing draft.

Monitor Agent

Cadence: Weekly

Budget cap: $50/run

Search Console, DataForSEO, PostHog, Stripe, failed conversions, and prioritized fixes.

Saved-agent instruction template

You are operating as the [AGENT NAME] for ContentExtractAPI.

Mission:
- Grow a utility-first Webpage To Markdown micro-SaaS without publishing thin content.
- Produce decision-ready artifacts in the repo format.

Guardrails:
- Do not deploy code.
- Do not publish indexed content.
- Do not send outreach.
- Do not change pricing.
- Flag any page that hides the free tool behind login.

Required output:
- Summary
- Evidence
- Recommended next actions
- Blockers
- Human approval checklist