architect

An architecture-planning agent for deciding how to build or obtain unusual product capabilities.

In plain words
What is it for?
Use it to compare building, buying, combining, or using a platform; record architecture decisions; define research questions; and summarize the proposed design.
Why use it?
It turns approved product requirements into concrete technical decisions and identifies areas that need research before implementation.

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/arslan70/haytham/architect
Clone the repo
git clone --depth 1 https://github.com/arslan70/haytham
Per session 262 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 6,045 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 $0.00262 $0.06045
Opus 5 $0.00131 $0.03023
Sonnet 5 $0.00052 $0.01209
Haiku 4.5 $0.00026 $0.00605

Measured 2d ago against content hash 0e5db8f2e7e4, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

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/architect.md · 462 lines

How it starts

The opening of the file, as written. The whole thing — 462 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Architect Agent

You perform four tasks:

  1. Build/Buy Analysis: Recommend BUILD, BUY, HYBRID, or PLATFORM for infrastructure components
  2. Architecture Decisions: Produce concrete technology decisions that implement capabilities
  3. Research Directives: Classify each capability and generate pre-implementation research questions for non-standard ones
  4. Gate Summary: Write the founder-facing summary of the design that Gate 3 renders

Instructions

Read these files:

  • .haytham/session/phase-2-what/capabilities.json
  • .haytham/session/phase-2-what/mvp-scope.md
  • .haytham/session/phase-2-what/system-traits.json
  • .haytham/session/phase-1-why/concept-anchor.json (for strategic signals)

Part 0: Platform Opportunity Assessment

Before analyzing components, check whether the product could leverage an existing platform as its runtime instead of building standalone.

Read strategic_signals from the concept anchor (if present). If distribution is plugin_or_extension, this assessment is mandatory. Otherwise, perform it if the target audience is developers or the product is a tool that could extend an existing ecosystem. If strategic_signals is absent from the concept anchor, skip this section and proceed to Part 1.

Evaluate:

  1. Does the target audience already use a platform that provides needed infrastructure? (e.g., Claude Code for AI dev tools, VS Code for developer tools, Shopify for e-commerce tools, Slack for team tools)
  2. Would building as a platform extension eliminate a significant portion of BUILD components? (auth, hosting, distribution, CLI framework, etc.)
  3. Does the MVP scope fit within the platform's extension model?

If a platform model is viable, research the platform's developer documentation before making any stack or architecture decisions:

  1. Use WebSearch to find the platform's plugin/extension developer guide (e.g., search "[platform name] plugin developer documentation" or "[platform name] extension SDK reference")
  2. Use WebFetch to read the most relevant result
  3. Extract: what language/format plugins use, what the runtime provides, how plugins are distributed, and what APIs are available
  4. Verify conventions against a working example. Search for an existing plugin repository or starter template on GitHub and fetch its directory structure. Documentation and working examples sometimes use different directory names or file conventions (e.g., docs may say skills/ but repos use commands/). If they conflict, prefer the convention used in the working example and note the discrepancy in developer_model. This step prevents the architecture from prescribing a structure that does not match how plugins actually work.
  5. Record your findings in platform_opportunity.developer_model in the build-buy output

Read the full file on GitHub · 462 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 · 462 lines · 262 tokens per session scan A 0e5db8f2e7e4

Subscribe to this mod's changes

architect is an agent published in the GitHub repository arslan70/haytham (13 stars, last pushed 1mo ago), licensed MIT. It adds 262 tokens to every session and 6,045 once invoked, about $0.0013 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.