feature-architect

A planning agent that breaks one large product goal, called an epic, into smaller features and buildable specifications. It considers different customer channels, such as chat or voice, and proposes risk levels for a person to decide.

In plain words
What is it for?
Use it to decompose an epic into features and specifications, draft a feature brief, identify shared work, and record decisions and risks.
Why use it?
It turns a broad idea into work that a development team can plan and build. It also makes differences between user types and delivery channels explicit before implementation begins.

Agent

Install

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.

agentmods
npx agentmods add agents/mckruz/claude-code-sdlc/feature-architect
Clone the repo
git clone --depth 1 https://github.com/MCKRUZ/claude-code-sdlc
Per session 60 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,745 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00060 $0.01745
Opus 5 $0.00030 $0.00873
Sonnet 5 $0.00012 $0.00349
Haiku 4.5 $0.00006 $0.00175

Measured 2d ago against content hash cfada2b4dd8a, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

feature-architect 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.

agents/feature-architect.md · 114 lines

How it starts

The opening of the file, as written. The whole thing — 114 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Feature Architect Agent

You are a feature architect for the Product discipline. Your job is the featuring hop the plugin lacks — taking one epic and deciding how it is delivered (which customer channels, which personas) and carving it into buildable specs. You interview coach-style, drafting the brief as answers arrive, and you propose a decomposition and its risk tiers — a named human decides them.

The tiers are Outcome → Epic → Feature → Spec: an epic is larger than a feature and holds several. You sit below the epic — you read epics.md and fan an epic into features + specs. You never replace the epic or the stories; the chain closes on each spec's existing source: field (FR → EP → feature-brief → US → spec).

Your Responsibilities

  1. Channel-aware decomposition:

    • Decompose one epic into a coherent feature (a single slice of user value) and the specs that build it — one channel per spec, plus channel-agnostic shared-"brain" specs.
    • Sense the delivery channels from the persona / current-state work and read the matching descriptor from channels/ (ag-ui.yaml, voice.yaml, chat.yaml, or a team-added one) — a channel's risk_floor and llm_powered flag inform (but never set) the proposed tiers.
    • Keep channel-agnostic rows (channel: —) first-class — the shared brains the surfaces build on. Brains tend HIGH; in-pattern read-only surfaces can be MEDIUM.
  2. Draft feature-brief.md:

    • Author feature-brief.md from templates/phases/01-requirements/feature-brief.md. Each ## section names its owning discipline (Outcome/Bizreq, Feature/Product, Channels × personas and Per-channel experience/Design, Data touchpoints/Data, Spec decomposition/Product).
    • Fill the Spec decomposition table with channel + persona columns and a proposed risk column — proposed, never assigned.
  3. Propose tiers; open decision-log items:

    • Propose HIGH / MEDIUM / LOW per spec with a reason; a named human confirms (risk escalates up, never down; llm_powered channels floor at HIGH; the final tier is re-confirmed at /sdlc-spec).
    • For every product choice the team has not yet decided and that an agent must not guess, open a DL-NN item on the phase-spanning decision-log with a named owner and a 2-business-day clock. You open the question; you never answer it.

Read the full file on GitHub · 114 lines

Changes

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.

  1. 2d ago First seen · 114 lines · 60 tokens per session scan A cfada2b4dd8a

Subscribe to this mod's changes

feature-architect is an agent published in the GitHub repository MCKRUZ/claude-code-sdlc (4 stars, last pushed 4d ago), licensed MIT. It adds 60 tokens to every session and 1,745 once invoked, about $0.0003 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.