hyperflow-amplify
Hyperflow prompt amplifier. Use when a prompt is rough, vague, or under-specified and should be rewritten to high quality before running it — "enhance this prompt", "make this prompt better", "improve my prompt", "rewrite this prompt". Domain-aware: detects the prompt's domain, injects the matching specialist standards + project rules, scores against an 8-dimension rubric, then offers to run it.
Allowed Tools
No tools specified
Provided by Plugin
hyperflow
Fifteen specialized slash commands turn one Claude session into a structured multi-agent engineering pipeline. Thinking models orchestrate, triage, and review; worker models execute in parallel — every step is a Worker → Reviewer pair, and every non-trivial phase fans into sub-phases with their own reviewers. Auto-routing is on by default — say 'audit the diff', 'debug this test', 'large migration', or 'run a workflow' and the orchestrator routes to the right skill without the /hyperflow:* prefix. /hyperflow:workflow uses Claude Code dynamic workflows for big tasks and a portable Codex/OpenCode adapter where native workflows are unavailable; /hyperflow:spec asks the questions a senior engineer would; /hyperflow:scope decomposes into a batched task graph; /hyperflow:dispatch fans out persona-stitched workers under tiered review; /hyperflow:amplify rewrites a rough prompt into a high-quality one before you run it. 15 composable personas, 6 adaptive flow profiles, and persistent project memory compound across sessions. Works across Codex App/CLI, Claude Code, OpenCode, and Antigravity.
Installation
This skill is included in the hyperflow plugin:
/plugin install hyperflow@claude-code-plugins-plusClick to copy
Instructions
hyperflow-amplify — prompt amplifier (Antigravity single-agent)
Turn a rough prompt into the single strongest version, then hand it off. Amplify never writes code — it produces a prompt other workflows (or you) execute. Follow the hyperflow doctrine.
Steps
- Read intent. Detect the prompt's domain(s) (frontend/ui, api/db backend, mobile, security, performance, refactor/bugfix/test, devops, docs). Read project rules when present:
AGENTS.md,~/.gemini/AGENTS.md,.hyperflow/memory/*. Note the gaps a senior engineer would have to guess. - Amplify. Rewrite into the single strongest version using this skeleton (adapt section presence — economy is a constraint, never inflate a one-liner into a spec):
- role / expertise framing (one line)
- precise task (unambiguous goal)
- Context (relevant background, current state, inputs)
- Constraints (domain doctrine standards + project rules — project rules win on conflict)
- Output (deliverable format + acceptance criteria)
- Out of scope (explicit anti-goals)
- Score the draft against the 8-dimension rubric (1-5 each): intent clarity · context sufficiency · scope boundaries · structure · domain doctrine injected · output spec · guardrails · economy. If any dimension < 4, revise once.
- Present the amplified prompt in one copy-ready fenced block + a 2-4 line rationale (what changed, which standards/project rules were injected, any flagged ambiguity).
- Handoff gate (AskUserQuestion, 4 options → mark one Recommended): Send to
hyperflow-spec(Recommended) · Send tohyperflow-scope· Send tohyperflow-dispatch· Copy only. On a "Send to…" choice, invoke that workflow with the amplified prompt.
Rules
- Never write code. Project rules override generic persona standards. No AI attribution in the amplified prompt or rationale.