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.
npx agentmods add rules/lukedj78/dev-flow/cursorrulesgit clone --depth 1 https://github.com/lukedj78/dev-flowWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00845 | $0.00845 |
| Opus 5 | $0.00423 | $0.00423 |
| Sonnet 5 | $0.00169 | $0.00169 |
| Haiku 4.5 | $0.00085 | $0.00085 |
Grade A, and why
cursorrules 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.
How it starts
The opening of the file, as written. The whole thing — 77 lines — stays where its author put it; the contents beside it link to each section on GitHub.
dev-flow — Cursor rules
You have access to dev-flow, a set of 8 reference documents (under dev-flow-skills/) that guide product development from idea to running app via a filesystem contract at .workflow/meta.json.
How to use
When the user's request matches one of these patterns, READ the corresponding skill's SKILL.md BEFORE doing the work:
| User says… | Read first |
|---|---|
| "I have an idea: …" / "I want to build X" | dev-flow-skills/dev-flow/SKILL.md (orchestrator) → routes to next |
| "Capture this idea / draft a PRD" | dev-flow-skills/prd-from-idea/SKILL.md |
| "Split the PRD into tasks" | dev-flow-skills/prd-to-tasks/SKILL.md |
| "Extract design from Figma" / "Use this Figma URL" | dev-flow-skills/figma-to-design-md/SKILL.md |
| "Use these images for the design" / "Make a DESIGN.md from these screenshots" | dev-flow-skills/image-to-design-md/SKILL.md |
| "Scaffold the app from DESIGN.md" / "Apply this DESIGN.md to shadcn/MUI" | dev-flow-skills/design-md-to-app/SKILL.md |
| "Build this page from a screenshot" | dev-flow-skills/screenshot-to-page/SKILL.md |
| "Add auth / db / payments / email / tests / CI" | dev-flow-skills/module-add/SKILL.md |
Read meta.json first
Before any non-trivial work, read <project-root>/.workflow/meta.json. The phase field tells you the current state. The stack field tells you which framework + UI library are in use. The artifacts field records what's been produced and from what.
If .workflow/ doesn't exist, run:
python3 dev-flow-skills/dev-flow/scripts/init_workflow.py . --name "<project>"
Tool name mapping
Skills use Claude Code naming. Cursor uses different tool names — translate semantically:
| Skill says | Cursor equivalent |
|---|---|
Bash |
terminal command |
Read |
filesystem read |
Edit (string replace) |
edit_file |
Write |
edit_file (new file) |
Glob |
find_files |
Grep |
grep_search |
State updates
After any skill produces a contract file (PROJECT.md, PRD.md, DESIGN.md, registry.json, generated pages, schema), record it:
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.
- yesterday First seen · 77 lines · 845 tokens per session scan A 6de151769a00
cursorrules is a cursor rule published in the GitHub repository lukedj78/dev-flow (5 stars, last pushed 3d ago), licensed MIT. It adds 845 tokens to every session, about $0.0042 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.
Other cursor rules, from other repositories
pi-en
PI Cognitive AI. Trigger: $pi/coding/dev/code/architecture/API/refactor/debug/bug/error/exception/crash/timeout/performance/optimization/test/compile/git/release/verify/review/CR/product/requirements/ops/growth/design/team/support, or deep/2+ failures/looping/stuck/giving-up/retry/nevermind.
minimax-mcp-tools
MCP and web-tool guidance: current-doc retrieval, direct-tool preference, MCP Apps structured content, and version-aware external lookups.
testing
Guide for writing pytest unit and integration tests.
best_practices
Advanced Best Practices covering Error Handling, Environment Variables, Performance, and Accessibility rules.
ai-actions-schemas
Cursor rule "ai-actions-schemas" from freshworks-developers/fw-dev-tools, covering ai actions schemas (actions.json), structure, parameters schema, preferred: flat structure and string constraints (enum, pattern, format).
tools
(none|read|write|network|exec) and risk (low|medium|high|critical). These drive the permission classifier — be honest.