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 instructions/jonradoff/lightcms/claude-mdgit clone --depth 1 https://github.com/jonradoff/lightcmsWhat 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 | $0.06115 | $0.06115 |
| Opus 5 | $0.03057 | $0.03057 |
| Sonnet 5 | $0.01223 | $0.01223 |
| Haiku 4.5 | $0.00611 | $0.00611 |
Grade A, and why
lightcms CLAUDE.md 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.
How it starts
The opening of the file, as written. The whole thing — 491 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Read VIBECTL.md for current project status, deployment details, and issue context before starting work.
LightCMS - Claude Code Memory
Project Overview
LightCMS is a lightweight, self-hosted content management system built in Go. It uses MongoDB for data storage and generates static HTML pages for public content serving.
Key URLs:
- Admin Dashboard:
/cm - Public Site:
/ - MCP Server:
bin/lightcms-mcp(stdio transport)
MCP Server Integration
IMPORTANT: All website content operations MUST go through the MCP server. Do NOT:
- Write scripts to directly modify the database
- Use Go code to create/edit/delete content
- Bypass the MCP server for any content management tasks
If the MCP server is not available or not working, ASK the user for permission before attempting any content changes through other means.
Starting the MCP Server
If the user requests a content operation and the MCP server is not yet running, start it using the convenience script:
./bin/lightcms-mcp-wrapper.sh
This wrapper script sets up the required environment variables (LIGHTCMS_CONFIG_DIR) and launches the MCP server. The MCP server uses stdio transport, so it will be connected automatically once started.
Content Operations (REQUIRE MCP or explicit user permission):
- Creating, editing, publishing, or deleting content
- Managing templates and their HTML layouts
- Uploading or managing assets (images, CSS, JS, documents)
- Updating theme settings (colors, fonts, header/footer HTML)
- Managing redirects, folders, and collections
- Viewing or reverting content versions
- Site configuration changes
Code Changes (allowed without MCP):
- Adding new features to LightCMS itself
- Fixing bugs in the application
- Changing application behavior or logic
- Adding new MCP tools
- Modifying database schemas or indexes
- Security improvements
⚠️ IMPORTANT: Rebuild MCP binary after adding tools
The Claude Desktop MCP integration uses a local binary (bin/lightcms-mcp) that connects to the remote server. The binary encodes all tool definitions — if you add or remove MCP tools without rebuilding it, Claude Desktop will see stale/missing tools.
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.
- 2d ago First seen · 491 lines · 6,115 tokens per session scan A a02060b9f3d2
lightcms CLAUDE.md is an instructions file published in the GitHub repository jonradoff/lightcms (27 stars, last pushed 1mo ago), licensed MIT. It adds 6,115 tokens to every session, about $0.0306 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 instructions, from other repositories
cctop AGENTS.md
Instructions for st0012/cctop, covering agents.md - development guide for cctop, scope and sources, required development rules, driver workflow and github and pr rules.
toh-framework CLAUDE.md
Instructions for wasintoh/toh-framework, covering claude.md — toh framework (repo development guide), what this is, everyday commands, verification protocol and single source, transformed per ide.
examples CLAUDE.md
Claude Code instructions for rossoctl/examples, covering claude.md - agent examples, repository structure, key commands, code style and dco sign-off (mandatory).
agent-rules child-process.instructions.md
Instructions for lirantal/agent-rules, covering system processes secure coding guidelines, your mission and spawning system processes.
claude-code-templates CLAUDE.md
Instructions for Justdvp/claude-code-templates, covering claude.md, project overview, development commands, package management and application commands.
grackle AGENTS.md
Instructions for nick-pape/grackle, covering grackle development guidelines, self-updating documentation, package readmes, environment variables and git workflow.