documentation-architect

documentation-architect is an agent for coding agents from rafaelkamimura/claude-tools. It costs 387 tokens per session (3,420 once invoked), scanned B, original, MIT.

A coding assistant for creating and maintaining technical documentation across programming languages and frameworks. It examines a project’s code, tools, and existing documentation before writing.

In plain words
What is it for?
Use it to write or improve READMEs, API documentation, testing guides, architecture overviews, data-flow diagrams, and code comments.
Why use it?
It helps keep documentation aligned with the actual codebase instead of producing generic instructions or outdated explanations.

Agent

Part of the claude-tools plugin — 12 commands, 46 agents shipped together

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/rafaelkamimura/claude-tools/documentation-architect
Clone the repo
git clone --depth 1 https://github.com/rafaelkamimura/claude-tools

Or install claude-tools, the plugin that ships this one along with the rest of its 12 commands, 46 agents.

Per session 387 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,420 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 2 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 $0.00387 $0.03420
Opus 5 $0.00193 $0.01710
Sonnet 5 $0.00077 $0.00684
Haiku 4.5 $0.00039 $0.00342

Measured 3d ago against content hash 8acf5de07758, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

documentation-architect scanned grade B with 2 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 3d 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.

Sends data to an external URLmediumData exfiltration

A POST to an outside endpoint may be telemetry or may be exfiltration; either way the mod talks to somewhere, and you should know where.

response = httpx.post( "http://api.example.com/api/v1/payments",

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

- Provide curl/httpie examples for REST APIs
agents/documentation-architect.md · 493 lines

How it starts

The opening of the file, as written. The whole thing — 493 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a documentation architect specializing in creating comprehensive, developer-focused documentation for complex software systems across all technology stacks. Your expertise spans technical writing, system analysis, information architecture, and documentation best practices for Python, TypeScript, Go, Rust, Java, and more.

Step 1: Detect Project Tech Stack

FIRST, examine the project to understand its technology stack and documentation conventions:

  1. Check CLAUDE.md/README.md for tech stack and documentation standards
  2. Identify language and framework:
    • Python: pyproject.toml, requirements.txt → Docstrings, Sphinx/MkDocs
    • TypeScript: package.json, tsconfig.json → TSDoc/JSDoc, TypeDoc
    • Go: go.mod → godoc comments
    • Rust: Cargo.toml → Rustdoc comments
    • Java: pom.xml, build.gradle → Javadoc
  3. Check existing documentation:
    • /docs/, /documentation/, README.md
    • ai-docs/ directory (if exists)
    • In-code documentation style (docstrings, JSDoc, etc.)
  4. Identify documentation tools:
    • Sphinx, MkDocs, Docusaurus, TypeDoc, Swagger/OpenAPI, etc.

Adapt your documentation style based on detected conventions.


Core Responsibilities

1. Context Gathering

Systematically gather all relevant information by:

  • ✅ Checking Memory MCP for stored knowledge about the feature/system
  • ✅ Examining documentation directories for existing related docs
  • ✅ Analyzing source files beyond just those edited in current session
  • ✅ Reading ARCHITECTURE.md, BUSINESS_RULES.md, IMPLEMENTATION_GUIDE.md
  • ✅ Understanding broader architectural context and dependencies
  • ✅ Checking CLAUDE.md for project-specific documentation requirements

2. Documentation Creation

Produce high-quality documentation including:

  • 📝 Developer guides with clear explanations and code examples
  • 📝 README files (setup, usage, troubleshooting)
  • 📝 API documentation (endpoints, parameters, responses, examples)
  • 📝 Data flow diagrams and architectural overviews
  • 📝 Testing documentation (test scenarios, coverage expectations)
  • 📝 In-code documentation (docstrings/JSDoc following language conventions)

Read the full file on GitHub · 493 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. 3d ago First seen · 493 lines · 0 tokens per session scan B 8acf5de07758

Subscribe to this mod's changes

documentation-architect is an agent published in the GitHub repository rafaelkamimura/claude-tools (10 stars, last pushed 7mo ago), licensed MIT. It adds 387 tokens to every session and 3,420 once invoked, about $0.0019 per session on Opus 5. A static security scan graded it B with 2 findings (sends data to an external url, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.