workflow

A set of required rules for how an AI coding agent works in a project, including when it may edit files, follow workflow commands, or commit changes.

In plain words
What is it for?
Enforcing proposal and confirmation gates, guiding task execution and acceptance reviews, and keeping agent-facing workflow rules consistent.
Why use it?
It prevents the agent from skipping review steps or making unapproved changes, while keeping the full instructions in the project's main workflow documents.

Cursor rule for Cursor

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/yuz9610/poe2overlay/workflow
Clone the repo
git clone --depth 1 https://github.com/yuz9610/POE2Overlay

Made for: Cursor.

Per session 653 This file is loaded in full into every session.
When invoked 653 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 $0.00653 $0.00653
Opus 5 $0.00327 $0.00327
Sonnet 5 $0.00131 $0.00131
Haiku 4.5 $0.00065 $0.00065

Measured 2d ago against content hash e458d8bb3c39, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

workflow 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 2d ago.

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.

.cursor/rules/workflow.mdc · 30 lines

How it starts

The opening of the file, as written. The whole thing — 30 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Workflow enforcement

Canonical (edit here first): docs/PROJECT_WORKFLOW.md § Agent 強制規則.
Prompts: docs/AI_AGENT_PROMPTS.md. Human cheat sheet: docs/AI_AGENT_SETUP.md. Docs ownership: docs/README.md. Workflow rules belong here in PROJECT_WORKFLOW.md; this rule is summary only.

When the user uses a workflow short command, follow the matching step in AI_AGENT_PROMPTS.md. Do not improvise a shorter path.

Five hard rules (summary only — do not expand here)

  1. First response = proposal only (read-only exploration OK). No writes, no commit, no Done.
  2. Skip gate only if the user says in the same message: 跳过确认 / 跳過確認 / skip confirmation / 直接修改 / 直接执行 / 直接實作 / 不用等确认 / 不必等確認. Vague urgency or In progress / Planned does not skip.
  3. 執行 task → Phase A execution plan first; Phase B after human confirms. In progress does not skip Phase A.
  4. Step 4 never commits. 驗收 task → acceptance review first; agent reads logs/*.log; commit only after human confirms Accept (plus log downgrade).
  5. Changing workflow rules → update PROJECT_WORKFLOW.md first, then sync AI_AGENT_PROMPTS.md intro, AGENTS.md, AI_AGENT_SETUP.md, this file.

Step 4 / Step 5 pointers

Step First reply After confirm
4 Execute Plan only Implement; fill task doc; no commit
5 Accept Accept / Needs changes / Blocked Done + commit per § Commit Message; Co-authored-by: CursorAgent; Windows git commit -F UTF-8

Step 3: new sprint → SPRINTS.md Current Sprint only (Planned). Step 6: SUMMARY.md per sprints/README.md § SUMMARY template; handoff IDs must exist in SPRINTS.md Task List; include workflow / project / docs review; update SPRINTS.md Current Snapshot, parent item status, Task List status / sprint fields, and reorder Above The Line Items / Task List with the same item order (in progress / started, then not started, then done); send larger improvements back to brainstorm / review / planning.

Read the full file on GitHub · 30 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. 2d ago First seen · 30 lines · 653 tokens per session scan A e458d8bb3c39

Subscribe to this mod's changes

workflow is a cursor rule published in the GitHub repository yuz9610/POE2Overlay (9 stars, last pushed 3mo ago), licensed MIT. It adds 653 tokens to every session, about $0.0033 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.