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/alphabitcore/nexus-gateway/sdd-workflowgit clone --depth 1 https://github.com/AlphaBitCore/nexus-gatewayWhat 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 | $0.01198 | $0.01198 |
| Opus 5 | $0.00599 | $0.00599 |
| Sonnet 5 | $0.00240 | $0.00240 |
| Haiku 4.5 | $0.00120 | $0.00120 |
Grade A, and why
sdd-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.
How it starts
The opening of the file, as written. The whole thing — 72 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Mandatory development workflow
Canonical text lives in
CLAUDE.md. This rule is the IDE-side surfacing; when this file disagrees withCLAUDE.md,CLAUDE.mdwins.
Forced defaults (every user request)
- Plan first — write a short plan (goal, approach, scope, risks, file touch list) before edits, multi-step work, or implementation. Use Plan Mode when non-trivial.
- Todo list always —
TodoWriteitems up front, mirroring the plan + the pipeline below. Keep statuses accurate (pending/in_progress/completed); never claim done with open or wrong-state todos. - Capture every user message as todos — every new request becomes a todo; new requests do NOT interrupt in-flight work, they queue (binding in CLAUDE.md).
- English only in the repository — write / edit only English in repo files (rules, docs, comments, UI strings, commits). Conversation language can match the user; files in the repo stay English.
- Deep reasoning — before substantive analysis or execution: clarify goals, read relevant code/docs, list assumptions, consider edge cases and failure modes. Don't rush.
- Ask when ambiguous — material choices not explicit in spec or chat: stop and ask with concrete options + recommended default. No silent guesses.
- Real implementation only — production / runtime code is fully implemented per spec. No
TODO/FIXME/ stubs / fake returns as the "solution". Mocks live in tests only. - No silent skips — skipping Plan, todos, or any pipeline step requires explicit user approval in chat.
- Commit reminder — when work is finished + verified, ask whether to commit. Never
git commitwithout explicit user instruction. - Never
git stash— parallel sessions share the working tree; a stash can destroy another session's work. Usego test -run Xorgit diff HEAD --instead.
Pre-edit reading (3-doc requirement)
See pre-edit-reading.mdc (binding). Before editing code you read:
- Architecture doc(s) via
docs/developers/architecture/README.md. - Feature doc(s) for any user-visible surface (
docs/users/features/cp-ui/,docs/users/features/agent-ui/,docs/users/features/flows/). docs/developers/workflow/conventions.mdfor code style.
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.
- 2d ago First seen · 72 lines · 1,198 tokens per session scan A 892579a57b8c
sdd-workflow is a cursor rule published in the GitHub repository AlphaBitCore/nexus-gateway (22 stars, last pushed 7d ago), licensed Apache-2.0. It adds 1,198 tokens to every session, about $0.0060 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 2026-08-30.
Other cursor rules, from other repositories
10-deterministic-audit
Enforce deterministic audit behavior — no invented controls, cite authoritative sources, emit traceable artifacts.
flagship_author
How to author a flagship in ks-cookbook.
recipe_author
How to author a recipe in ks-cookbook.
30-framework-routing
Route tasks to HIPAA, HITECH, PCI-DSS, SOC 2, ISO 27001, NIST CSF, NIST AI RMF, FERPA, COPPA, CCPA, US state privacy, GDPR, FedRAMP, SOX, CMMC, and GLBA skills based on user intent and presets.
00-compliance-agent-core
Core compliance agent mission, PHI redaction gate, and audit lifecycle for HIPAA, PCI-DSS, and SOC 2.
20-phi-redaction-first
PHI redaction first — Presidio gate, token handling, and minimum necessary LLM access for HIPAA workflows.