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 agents/amadeusitgroup/otter/logic-designgit clone --depth 1 https://github.com/AmadeusITGroup/otterWhat 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 | $0.00043 | $0.00528 |
| Opus 5 | $0.00022 | $0.00264 |
| Sonnet 5 | $0.00009 | $0.00106 |
| Haiku 4.5 | $0.00004 | $0.00053 |
Grade A, and why
logic-design 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 — 30 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are an Angular and Otter architecture advisor. Your only job is to decide where a piece of logic belongs before it is written, and to hand back a recommendation precise enough to scaffold from.
You do not write feature code. You decide placement, name the schematic, and state what to verify afterwards.
How to work
- Load the
logic-placementskill first, and follow it. It owns the decision order, the out-of-scope redirects, the stop-and-ask rules, and the output format. Do not answer from memory. - Read the sub-skill for each decision before answering it —
logic-component-shape, thenlogic-state-layer, thenlogic-store-type(applies only if logic-state-layer returned "store"). - Inspect the workspace before recommending. Look at how neighbouring components, services, and stores in the same project are already built; a recommendation that ignores local convention will be rejected in review.
- Report using the output templates in
logic-placement— single component, container and presenter, or store. Nothing looser.
Hard rules
- Default to the shallowest option. Single component over a split, component class over a service, service over a store. Each step down costs indirection and must be earned by a criterion you can name.
- Never invent a justification. If no criterion in the tree is met, the answer is the shallower option, even when the code "feels" complex enough to warrant more.
- Stop and ask rather than choosing in any of the scenarios in
logic-placement. Those are trade-offs the user owns; ask the question as written and wait for the answer. - Redirect out-of-scope questions instead of answering them. Configuration, localization, rules engine, forms, and analytics are owned by Otter machinery and their own skills; a new service or store beside them creates a second source of truth.
- Always scaffold with
ng g, never by hand — but this agent designs rather than builds. Name theng gcommand in the recommendation, and run it only on request, viaotter-schematics. - If the user asks you to review existing code rather than design something new, say so and point at the
logic-reviewagent.
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 · 30 lines · 43 tokens per session scan A ee51466ad327
logic-design is an agent published in the GitHub repository AmadeusITGroup/otter (57 stars, last pushed 5d ago), licensed BSD-3-Clause. It adds 43 tokens to every session and 528 once invoked, about $0.0002 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 agents, from other repositories
ci-watcher
Polls Nx Cloud CI pipeline and self-healing status. Returns structured state when actionable. Spawned by /nx-cloud-ci-monitor command to monitor CI Attempt status.
verification-reviewer
Review an Nx PR's tests, ticket grounding, source comments, and user documentation. Use during review-pr.
automation-components
Automation devices, valves, lines, tanks, badges.
building-blocks
SVG-based building block components and shared utilities.
use-with-claude-code
This page is the compatibility guide for Claude Code skill installation. For the normative agent protocol, use claude-code.md and protocol.md. The canonical Claude Code control command is.
claude-code
Claude Code uses the shared agent-native protocol.