Agent skill

Architect Fast

by DanMcInerney in DanMcInerney/architect-loop

A skill your agent uses when the user asks to architect-fast a change, run the light factory lane, or factory-build a small goal — a few files, roughly one sitting, at most ~3 parallel issues — into…

MITAuto-check passedAgent Workflows

Install Architect Fast

skills CLI
$ npx skills add DanMcInerney/architect-loop --skill architect-fast -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install DanMcInerney/architect-loop architect-fast --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/DanMcInerney/architect-loop.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/architect-fast .claude/skills/architect-fast && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
architect-fast
GitHub stars
626
Token cost
~1.9k tokens
SKILL.md length
898 words
Files
1
Skills in repo
11
Repo updated
First seen
Licence
MIT

At a glance

A skill your agent uses when the user asks to architect-fast a change, run the light factory lane, or factory-build a small goal — a few files, roughly one sitting, at most ~3 parallel issues — into…

  • Works in 7 steps: Ground. Load codebase-design first and… → Intake. No question batch; anything… → Decompose. to-issues cuts at most 3… → …
  • The user asks to architect-fast a change
  • SKILL.md covers Rules kept, Relaxed by design, Procedure and Substitutions, plus 1 more section
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Architect Fast is an agent skill from DanMcInerney/architect-loop. Use when the user asks to architect-fast a change, run the light factory lane, or factory-build a small goal — a few files, roughly one sitting, at most ~3 parallel issues — into a specced plan with parallel builders, one builder-performed closing review-and-fix, and a single PR. Same loop shape as /architect with the per-issue grading machinery removed; for larger, riskier, or many-issue work use /architect instead.

Its SKILL.md is about 1.9k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Agent Workflows. The repository describes itself as: Super optimized /goal loop. Massive token savings and higher quality. Smart model designs and reviews, cheaper model builds.. The licence is MIT.

When your agent uses it

  • The user asks to architect-fast a change
  • Run the light factory lane
  • Factory-build a small goal — a few files
  • Roughly one sitting

Example prompts

  • “/architect-fast”

Workflow steps

7 steps, taken from the first numbered list in SKILL.md.

  1. Ground. Load codebase-design first and use its glossary exactly.
  2. Intake. No question batch; anything genuinely open becomes a timed
  3. Decompose. to-issues cuts at most 3 tracer-bullet vertical-slice
  4. Factory loop. Dispatch every ready issue as a fresh
  5. Closing test pass. After all issues merge, run the full test suite
  6. Builder review. Dispatch one fresh worktree-isolated builder with
  7. Integrate. Dispatch one subagent running the integrate stage

What it can do on your machine

Read from SKILL.md and the folder at commit 28dca7d. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Architect Fast loads about 1.9k tokens when it runs. Until then it costs about 109 tokens; SKILL.md has 898 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~109
When it runs · the whole SKILL.md, loaded when a task matches
~1.9k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from DanMcInerney/architect-loop at commit 28dca7d, republished under its MIT licence (© DanMcInerney). 898 words, ~1,904 tokens.

Download SKILL.mdSave it as .claude/skills/architect-fast/SKILL.md (or your agent's skills folder).
name
architect-fast
description
Use when the user asks to architect-fast a change, run the light factory lane, or factory-build a small goal — a few files, roughly one sitting, at most ~3 parallel issues — into a specced plan with parallel builders, one builder-performed closing review-and-fix, and a single PR. Same loop shape as /architect with the per-issue grading machinery removed; for larger, riskier, or many-issue work use /architect instead.
effort
high

Architect Fast

You are the orchestrator for the light factory lane: /architect's loop shape — spec, parallelizable issues, fresh isolated builders, one closing PR — with the per-issue grading machinery removed. Reuse /architect's machinery by pointer, never by duplication: skills/architect/dispatch.md (## Model alias table, ## Builder block template, ## Preflight and postflight dispatch, ## Status display) and skills/architect/tracker.md (## Config) govern models, dispatch, merges, status, and tracker modes, except where a substitution below names a deviation.

Size ceiling: at most 3 builder issues and ~400 changed lines expected. If the honest decomposition needs more, stop and recommend /architect — a hard stop, human ruling, never silently absorbed.

Rules kept

Unchanged from /architect: the tracking issue is the coordination log; builders are fresh, worktree-isolated, and never commit; the orchestrator owns commits, merges, and closure; tier is set at decomposition, never moved by a failure; the timed-ruling protocol times every human question and no approval gate exists anywhere; execution-conflict checks are mandatory at PHASE 0; no silent fallback — every precondition, blocker, and substitution is recorded.

Relaxed by design

  • No strategist subagents — no adversarial spec, decomposition, or final review. The orchestrator writes the spec itself and to-issues runs straight off it.
  • No frozen check files, no check-runner: issue-body acceptance criteria, builder-run tests, and the closing builder review carry the weight.
  • No watchdog script — a per-wave timed background sleep is the stall-fallback wake.
  • "Nobody grades their own work" (Hard Rules 3 and 4) is deliberately relaxed: the closing review is one fresh builder subagent that both reviews and fixes, with no check-runner over its work — the orchestrator's merge through postflight and the closing PR are the later eyes. Product docs land in integrate's docs pass, same as /architect.

Procedure

  1. Ground. Load codebase-design first and use its glossary exactly. Read authority docs in /architect's order. Tracker preflight and the backend canary are unchanged (skills/architect/tracker.md ## Preflight per mode). No ground script mid-run — its reconcile assumes checkrun evidence this lane never produces; read the ≤3-issue frontier from tracker state. skills/architect/status.ps1|.sh works unchanged; the manifest sets lane: architect-fast so status renders the light-lane graph.

  2. Intake. No question batch; anything genuinely open becomes a timed ruling (skills/architect/SKILL.md ## Timed-ruling protocol) recorded as an assumption. No adversarial spec review. Write the spec with to-spec on its exact template, one named substitution: ## Validation strategy names this lane's gates — builder-run tests plus the closing builder review-and-fix — never the check-runner or closing review. Create the tracking issue and manifest; cut the factory branch. Append status breadcrumbs to docs/runs/<run>/status-events.jsonl with the same JSONL contract as /architect, using fast-lane stages where they differ: publish, issues, dispatch-head, timed-wake, and closing-test.

  3. Decompose. to-issues cuts at most 3 tracer-bullet vertical-slice issues with a file-disjoint frontier and producer interface contracts, two named substitutions: an acceptance-criteria section replaces the check path, and the change-skeleton is optional. Above the ceiling, stop and recommend /architect.

  4. Factory loop. Dispatch every ready issue as a fresh worktree-isolated builder — same agent def, preloads, model resolution, never-commit rule, PHASE-0 execution conflicts, raw-evidence reports (skills/architect/dispatch.md ## Builder block template), one substitution: the frozen-checks section becomes acceptance criteria quoted from the issue body. At dispatch, record each job's dispatch-head SHA on its issue — the job's ffcheck target and postflight base (freeze_sha field, same semantics). Each wave arms one timed background sleep as the stall-fallback wake; on a fallback wake with jobs in flight, judge liveness from report growth directly. On DONE, merge through postflight (skills/architect/dispatch.md ## Preflight and postflight dispatch). On BLOCKED, answer durably and respawn fresh.

  5. Closing test pass. After all issues merge, run the full test suite over the merged factory branch and capture the raw output.

  6. Builder review. Dispatch one fresh worktree-isolated builder with the run diff scope, the test-pass output, and /architect's calibration wording: "Flag only gaps that affect correctness, the stated requirements, or documented project invariants -- cite file:line evidence for every finding. Do not report stylistic preferences." It reviews code correctness, cross-slice cohesion, and test stewardship; records its findings as one fix issue (orchestrator-filed where the sandbox has no tracker access); then fixes them in its worktree. The orchestrator merges through postflight and closes the issue — builders still never commit. This is the lane's only review; post the verdict plus diffstat on the tracking issue.

  7. Integrate. Dispatch one subagent running the integrate stage skill (skills/integrate/SKILL.md) with the change-context digest; its first step is the docs pass. Standing ruling: the builder-review verdict is the recorded final-review substitute — integrate's "skipped by a recorded ruling" arm — and its graded-RUN verification set is empty by design; validator-suite verification still runs. The orchestrator rules on the result and posts the digest.

Show full SKILL.md (139 more words)Show less

Stage skills invoked explicitly: codebase-design, to-spec, to-issues, integrate. frozen-checks, adversarial-review, and final-review are intentionally not invoked in this lane.

Substitutions

Reused contractFast-lane substitution
to-spec's validation-strategy namingBuilder-run tests + the closing builder review-and-fix
to-issues' per-issue check pathAcceptance-criteria section in the issue body
to-issues' mandatory change-skeletonOptional
Builder template's frozen-checks sectionAcceptance criteria quoted from the issue body
Postflight's freeze_sha diff baseThe job's recorded dispatch-head SHA
Ground script's frontier lineRead from tracker state; no ground script mid-run
WatchdogOne per-wave timed background sleep
Final reviewOne fresh builder review-and-fix subagent fed the test-pass output; fixes merge through postflight
Integrate's graded-RUN verification setEmpty by design; validator-suite verification still runs

Maintenance

Re-read against each new model generation; the size ceiling and Rules kept are invariants, the rest is prunable. Re-run docs/evals/trigger-prompts.md. No feature ships without evidence in DESIGN.md.

© DanMcInerney, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in skills/architect-fast of DanMcInerney/architect-loop.

Open the folder on GitHubat commit 28dca7d

Compare with similar skills

Architect Fast next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Architect Fast compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Architect Fast this skillDanMcInerney/architect-loop626—~1.9kAutomated safety check: PassMIT
MCP Server Builderanthropics/skills180k63 repos~2.3kAutomated safety check: PassApache-2.0
Hook Development for Claude Code Pluginsanthropics/claude-plugins-official38k10 repos~4.1kAutomated safety check: NotesApache-2.0
Using Superpowersfarm-fe/farm5.6k35 repos~1.4kAutomated safety check: PassMIT
Executing Plans Inlineobra/superpowers297k2 repos~5.1kAutomated safety check: PassMIT
Skill CreatorAzure/azqr79589 repos~8.2kAutomated safety check: PassApache-2.0

Similar skills

  • MCP Server Builder

    anthropics/skills

    Official

    Guides the design and implementation of Model Context Protocol servers in TypeScript or Python, from tool naming and error messages to evaluation.

    180k GitHub starsUsed in 63 repos~2.3k tokens
    Agent WorkflowsAuto-check passed
  • Hook Development for Claude Code Plugins

    anthropics/claude-plugins-official

    Official

    Explains how to write Claude Code plugin hooks, both prompt-based checks and bash commands, for events such as PreToolUse, Stop and SessionStart.

    38k GitHub starsUsed in 10 repos~4.1k tokens
    Agent WorkflowsAuto-check: notes
  • Using Superpowers

    farm-fe/farm

    A skill your agent uses when starting any conversation - establishes how to find and use skills, requiring Skill tool invocation before ANY response including clarifying questions

    5.6k GitHub starsUsed in 35 repos~1.4k tokens
    Agent WorkflowsAuto-check passed
  • Executing Plans Inline

    obra/superpowers

    Has the agent carry out an implementation plan itself, task by task in the current session, keeping a ledger, proving each step with a test and ending with one whole-branch review.

    297k GitHub starsUsed in 2 repos~5.1k tokens
    Agent WorkflowsAuto-check passed
  • Skill Creator

    Azure/azqr

    Official

    Create new skills, modify and improve existing skills, and measure skill performance.

    795 GitHub starsUsed in 89 repos~8.2k tokens
    Agent WorkflowsAuto-check passed
  • Claude Code Agent Development

    anthropics/claude-plugins-official

    Official

    Explains how to write agents for Claude Code plugins: the markdown file with YAML frontmatter, trigger descriptions, model and color settings, and system prompt design.

    38k GitHub starsUsed in 7 repos~2.8k tokens
    Agent WorkflowsAuto-check passed

More from DanMcInerney/architect-loop

All 11 skills in this repo
  • Architect Research

    DanMcInerney/architect-loop

    A skill your agent uses when the user asks for discovery-scale research that informs a decision: brainstorming a project or feature, choosing a technology, or requests like "research X", "what's the…

    626 GitHub stars~2.3k tokensUpdated 25 days ago
    Auto-check passed
  • Adversarial Review

    DanMcInerney/architect-loop

    A skill your agent uses when the architect factory orchestrator dispatches a fresh strategist subagent to harden a draft spec: falsify it with file:line evidence, fold the surviving findings into a…

    626 GitHub stars~1.1k tokensUpdated 25 days ago
    Auto-check passed
  • Architect

    DanMcInerney/architect-loop

    A skill your agent uses when the user asks to architect, run or continue the autonomous software factory, turn a goal into a hardened tracker issue plan, dispatch builder jobs, grade finished work…

    626 GitHub stars~2.5k tokensUpdated 25 days ago
    Auto-check passed
  • Final Review

    DanMcInerney/architect-loop

    A skill your agent uses for the closing whole-run review in the architect factory — the only model review in the loop: dispatched by the orchestrator, at finish, to one fresh strategist subagent…

    626 GitHub stars~1.7k tokensUpdated 25 days ago
    Auto-check passed
  • Frozen Checks

    DanMcInerney/architect-loop

    A skill your agent uses when the strategist drafts per-issue graded checks after decomposition and before builder dispatch.

    626 GitHub stars~816 tokensUpdated 25 days ago
    Auto-check passed
  • TDD

    DanMcInerney/architect-loop

    Test-driven development for factory builders. An agent skill from DanMcInerney/architect-loop.

    626 GitHub stars~730 tokensUpdated 25 days ago
    Auto-check passed

Categories

Questions about Architect Fast

What does Architect Fast do?

A skill your agent uses when the user asks to architect-fast a change, run the light factory lane, or factory-build a small goal — a few files, roughly one sitting, at most ~3 parallel issues — into…. Architect Fast is an agent skill from DanMcInerney/architect-loop. Use when the user asks to architect-fast a change, run the light factory lane, or factory-build a small goal — a few files, roughly one sitting, at most ~3 parallel issues — into a specced plan with parallel builders, one builder-performed closing review-and-fix, and a single PR.

When should I use Architect Fast?

Architect Fast fits situations like: the user asks to architect-fast a change; run the light factory lane; factory-build a small goal — a few files; roughly one sitting.

How do I install Architect Fast in Claude Code?

Run `npx skills add DanMcInerney/architect-loop --skill architect-fast -a claude-code`. Or copy the skill folder (skills/architect-fast in DanMcInerney/architect-loop) into .claude/skills/architect-fast in your project. Claude Code loads it when a task matches its description.

How do I install Architect Fast in Codex?

Run `npx skills add DanMcInerney/architect-loop --skill architect-fast -a codex`. Or copy the skill folder (skills/architect-fast in DanMcInerney/architect-loop) into .agents/skills/architect-fast in your project. Codex loads it when a task matches its description.

Can I use Architect Fast in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add DanMcInerney/architect-loop --skill architect-fast -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/architect-fast, .gemini/skills/architect-fast, .github/skills/architect-fast and .opencode/skills/architect-fast in your project.

What does Architect Fast need to run?

SKILL.md names no scripts, command-line tools or credentials: Architect Fast is instructions for the agent only.

Does Architect Fast access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Architect Fast safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Architect Fast use?

Architect Fast is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Architect Fast use?

About 1.9k tokens (SKILL.md is roughly 7.6k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Architect Fast?

Skills that share tags, products or a category with Architect Fast: MCP Server Builder (anthropics/skills, 180k stars), Hook Development for Claude Code Plugins (anthropics/claude-plugins-official, 38k stars), Using Superpowers (farm-fe/farm, 5.6k stars) and Executing Plans Inline (obra/superpowers, 297k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Architect Fast?

DanMcInerney (a GitHub user) maintains it in DanMcInerney/architect-loop, which has 626 GitHub stars. The repository holds 11 skills in this directory. The repository was last updated on September 13, 2026.

Source: DanMcInerney/architect-loop on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.