Dirac is an open-source coding agent for carrying out complex software-engineering tasks in editors, terminals, and Agent Client Protocol clients. It is used by developers who want autonomous task execution, precise codebase editing, parallel work, and control over model providers.
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 skills add dirac-run/dirac --skill architecture-boardgit clone --depth 1 https://github.com/dirac-run/diracWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/dirac-run/dirac/architecture-board)<a href="https://agentmods.dev/skills/dirac-run/dirac/architecture-board"><img src="https://agentmods.dev/badge/skills/dirac-run/dirac/architecture-board/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/dirac-run/dirac/architecture-board"><img src="https://agentmods.dev/badge/skills/dirac-run/dirac/architecture-board.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00027 | $0.06549 |
| Opus 5.5 | $0.00011 | $0.02620 |
| Sonnet 5.5 | $0.00005 | $0.01310 |
| Haiku 4.5 | $0.00003 | $0.00655 |
Grade A, and why
architecture-board 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 4d 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 — 219 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Architecture Board
Use architecture_diagram to publish and recover semantic architecture diagrams in this foreground Task or Goal conversation. Explore the code with existing tools; the diagram tool does not establish source-code facts. Do not activate this skill in private contained tasks, subagents, or utility models, or inject saved diagrams into their context automatically.
Intent-driven defaults
Infer the user's intent from the conversation; do not ask them to choose a diagram type, abstraction level, or mode. Choose a useful scope and state the view briefly in the title/summary, e.g. "Runtime ownership and main execution path." Ask only when an unresolved requirement or baseline materially changes the answer and cannot reasonably be inferred. Explicit user scope takes precedence over these defaults.
| Intent | Default view | Essential answer |
|---|---|---|
| Understand something | Current ownership boundaries and the main execution path | What are the pieces, who owns them, how do they cooperate, and what is shared? |
| Design / brainstorm / refine | One recommended proposed architecture with constraints and open decisions | What are we proposing, why, and which decisions still matter? |
| Review a proposal or code change | Comparable before/after views of the affected architecture | What changes in responsibility, ownership, dependencies or behavior, and what stays the same? |
Understand: mental model before inventory
- Start with a small overview, usually 6–12 meaningful nodes; this is a guideline, not a quota. Preserve essential boundaries rather than forcing a node count. Show ownership and the principal path, not every class and call.
- Make important instance counts, shared resources and execution boundaries visible in short node/edge labels, such as "one per worker," "shared workspace," "owns" and "await all." Descriptions add evidence and nuance; they must not repair a misleading canvas.
- Keep deep execution, failure/recovery and observability detail out of the overview unless they are the question. "In depth" means focused explanation, not automatic graph inflation: preserve the overview and publish a separately named detail board for the relevant area. Each successful save automatically opens its board. Identify the overview/detail relationship by diagram name in summaries; do not invent clickable graph links or schema fields.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 4d ago First seen · 219 lines · 27 tokens per session scan A 39452a883dca
architecture-board is a skill published in the GitHub repository dirac-run/dirac (1,541 stars, last pushed yesterday), licensed Apache-2.0. It adds 27 tokens to every session and 6,549 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-10-23.
Other skills, from other repositories
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…