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 agentmods add commands/benkapner/claude-code-basecamp/architecture-docsgit clone --depth 1 https://github.com/Benkapner/claude-code-basecampWhat 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 | $0.00028 | $0.00948 |
| Opus 5 | $0.00014 | $0.00474 |
| Sonnet 5 | $0.00006 | $0.00190 |
| Haiku 4.5 | $0.00003 | $0.00095 |
Grade A, and why
architecture-docs 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 yesterday.
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 — 124 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Architecture Documentation
Generate architecture documentation for the project.
Quick Mode
If $ARGUMENTS contains --quick or quick:
-
Read the project — README, entry points, directory structure. Understand what the system does and how data/control flows through it.
-
Write a Mermaid flowchart showing the main flow. Keep it simple:
- Use
flowchart LR(left-to-right) for pipelines,flowchart TD(top-down) if it fits better - 5-12 nodes max — major components only, not individual files
- Label arrows with what moves between components (data, events, etc.)
- Use subgraphs sparingly — only if there's a clear grouping
- Name nodes by what they do, not filenames
- Use
-
Save as a self-contained HTML file that renders the diagram using Mermaid CDN:
uv run .ai-workspace/scripts/mktmpdir.py visualize 2>/dev/null || mkdir -p .tmp/visualizeSave to
.tmp/visualize/architecture.html. The HTML is just a minimal page that loadshttps://cdn.jsdelivr.net/npm/mermaid/dist/mermaid.min.jsand renders the diagram — no extra UI. -
Show the user the Mermaid source inline and tell them the file path.
Stop here for quick mode — do not produce full documentation.
Full Mode (default)
Step 1: Discover Architecture
Use parallel agents to explore the codebase:
- Read README.md, CLAUDE.md, and any existing docs
- Map the directory structure and identify key modules
- Trace data flow: where does data enter, how is it processed, where does it go?
- Identify external integrations (APIs, databases, services)
- Review config files for infrastructure context
Step 2: Generate Documentation
Produce the following sections, scaled to the project's complexity:
System Overview
- What the system does (1-2 paragraphs)
- Key stakeholders and users
- External systems it connects to
Component Architecture
- Mermaid diagram showing major components and their relationships
- Each component: responsibility, inputs, outputs, dependencies
- Clear boundaries between modules
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.
- yesterday First seen · 124 lines · 28 tokens per session scan A 2d6c763451d2
architecture-docs is a command published in the GitHub repository Benkapner/claude-code-basecamp (16 stars, last pushed 12d ago), licensed MIT. It adds 28 tokens to every session and 948 once invoked, about $0.0001 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.
Other commands, from other repositories
spec-forge
Use when generating software specifications — full chain (Idea→Decompose→Tech Design + Feature Specs) or individual documents.
propagate
Use after editing an upstream doc (PRD/SRS/tech-design/feature-spec) to propagate changes downstream and keep the entire doc chain consistent.
review
Use when reviewing spec-forge generated documents for quality, completeness, and consistency — auto-fixes issues if found.
analyze
Use when analyzing a document collection to map themes, find conflicts, gaps, and redundancies — generates landscape analysis report.
audit
Use when auditing existing project docs for quality, completeness, and code alignment — generates findings report with fix recommendations.
test-cases
Use when writing test cases, generating tests, supplementing test coverage, or improving test completeness — auto-scans project, designs multi-dimensional test cases with coverage matrix.