prd-pr-cursor

A set of Cursor-specific rules for running the prd-pr feature-planning workflow, including the correct names and models for its task subagents. A subagent is a separate AI worker assigned to a particular development phase.

In plain words
What is it for?
Use it when running prd-pr in Cursor or implementing work from its task plan, especially when dispatching exploration, planning, backend, frontend, and test tasks.
Why use it?
It prevents Cursor from using Claude Code's task syntax or model labels incorrectly. It maps exploration, architecture, review, implementation, and testing phases to the expected subagent types.

Cursor rule

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add rules/wei6bin/skills/prd-pr-cursor
Clone the repo
git clone --depth 1 https://github.com/wei6bin/skills
Per session 1,463 This file is loaded in full into every session.
When invoked 1,463 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.01463 $0.01463
Opus 5 $0.00732 $0.00732
Sonnet 5 $0.00293 $0.00293
Haiku 4.5 $0.00146 $0.00146

Measured yesterday against content hash 742aa85c97d7, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

prd-pr-cursor scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

prd-pr-cursor/rules/prd-pr-cursor.mdc · 94 lines

How it starts

The opening of the file, as written. The whole thing — 94 lines — stays where its author put it; the contents beside it link to each section on GitHub.

prd-pr on Cursor (adapter)

The prd-pr plugin targets Claude Code (agent_type: "prd-pr:…" and model: in agents/*.md). In Cursor, use this rule whenever running the orchestrator skill or implementing from docs/new-feature/*/04-task-plan.md.

Dispatch (Task tool)

Use subagent_type — never agent_type, never the prd-pr: prefix.

Phase subagent_type Notes
2 — exploration code-explorer 2–3 parallel tasks, different angles
4 — architecture code-architect One task with Phase 2–3 context
6 — plan review plan-reviewer Two parallel tasks (AC/security vs tasks/tests)
8 — backend half impl-backend Scope e.g. "SLICE-01 backend half"; dispatched concurrently with the frontend half against the slice's frozen Contract:. Append the lean mode from the slice's story-point size — lean: lite (1–2 pts) / lean: full (3+, default). Large slices (8+/spike) stay full — the implementer flags over-scope, never drops an AC
8 — frontend half impl-frontend Dispatched concurrently with the backend half (single batch), scope "SLICE-01 frontend half — build against the frozen contract with a typed mock; do not wait for the backend; stay on the mock, no per-slice integration"; append the same lean: token. One "whole-story integration" re-dispatch at the end (after every slice lands) swaps every slice's mock for the real backends and reconciles drift across the story
8 — simplify impl-simplify Per slice, changed files only
9 — walkthrough test-plan-walker Clean context; do not inline Phase 9 in parent

Do not substitute generic explore or generalPurpose for the rows above unless the user explicitly asks for a one-off search outside prd-pr.

Models

Project agents: .cursor/agents/{name}.md — Cursor reads model: frontmatter here (not plugin sonnet/opus/haiku). See .cursor/agents/README.md.

Subagent .cursor/agents model:
code-explorer composer-2
code-architect claude-opus-4-7
plan-reviewer gpt-5.5
impl-backend, impl-frontend composer-2.5
impl-simplify composer-2.5-fast
test-plan-walker inherit

Read the full file on GitHub · 94 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 94 lines · 1,463 tokens per session scan A 742aa85c97d7

Subscribe to this mod's changes

prd-pr-cursor is a cursor rule published in the GitHub repository wei6bin/skills (2 stars, last pushed 9d ago), licensed MIT. It adds 1,463 tokens to every session, about $0.0073 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.