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/dejuknow/md-redline/agents-mdgit clone --depth 1 https://github.com/dejuknow/md-redlineWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/dejuknow/md-redline/agents-md)<a href="https://agentmods.dev/instructions/dejuknow/md-redline/agents-md"><img src="https://agentmods.dev/badge/instructions/dejuknow/md-redline/agents-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.29130 | $0.29130 |
| Opus 5 | $0.14565 | $0.14565 |
| Sonnet 5 | $0.05826 | $0.05826 |
| Haiku 4.5 | $0.02913 | $0.02913 |
Grade A, and why
md-redline AGENTS.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 5d 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 — 1,881 lines — stays where its author put it; the contents beside it link to each section on GitHub.
md-redline
Use README.md as the canonical product, usage, and feature reference.
This is the single source of truth for agent-facing docs. CLAUDE.md contains only Claude-specific skill routing and points here. Do not duplicate content between the two files.
Repo snapshot
- Local web app for inline review comments in markdown files
- Frontend: React 19 + TypeScript + Tailwind CSS v4 + Vite 8
- Backend: Hono server in
server/index.ts - CLI entry:
bin/md-redline, exposed asmdrandmd-redline; the body isbin/cli.js
Architecture overview
A Hono server serves both a Vite-built React SPA and a REST API. An optional MCP stdio server lets AI agents request human review and wait for feedback.
Directory layout
bin/md-redline— published entry point (mdrcommand), a shim that importsbin/cli.jsbin/cli.js— the CLI itself. Starts server, opens browser, handlesmdr mcp install.server/— Hono app, routes, review session store, MCP stdio layersrc/— React SPA (Vite + TypeScript + Tailwind)src/lib/comment-parser.ts— core parser for inline comment markerssrc/lib/agent-prompts.ts— prompt builder for agent handoffe2e/— Playwright E2E testseval/— eval harness and fixtures
Behavior that matters
- Comments are stored inline as
<!-- @comment{...} -->markers immediately before their anchor text. - Default workflow: comments are instructions to an agent, and addressed markers should be removed.
- Optional resolve workflow adds
open/resolvedstatus for human review. MarkdownViewer.tsxowns highlight DOM imperatively via refs; React does not reconcile the rendered markdown subtree.- Overlapping comments are allowed and depend on marker position plus stored anchor/context for rematching.
- URL boot params: a
?file=boot opens that file active on top of the restored tab session, and?dir=re-roots the explorer with tabs restored as saved; only?review=skips session restore.
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.
- 5d ago First seen · 1,881 lines · 29,130 tokens per session scan A 854ef6d60b11
md-redline AGENTS.md is an instructions file published in the GitHub repository dejuknow/md-redline (35 stars, last pushed 4d ago), licensed MIT. It adds 29,130 tokens to every session, about $0.1457 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
onebrain GEMINI.md
Instructions for onebrain-ai/onebrain, covering onebrain — gemini cli, load order, project-level gemini config, tool name mapping and agent instructions.
onebrain AGENTS.md
Instructions for onebrain-ai/onebrain, covering onebrain — codex cli, load order, tool name mapping and agent instructions.
onebrain CLAUDE.md
Instructions for onebrain-ai/onebrain, covering onebrain — claude code and agent instructions.
memory-harbor AGENTS.md
AGENTS.md instructions for qiontoo/memory-harbor, covering memory harbor instructions, system identity, startup, placement and safety.
Physical-AI-Humanoid-Robotics-Book GEMINI.md
Gemini CLI instructions for DevExplorerr/Physical-AI-Humanoid-Robotics-Book, covering gemini cli rules, task context, core guarantees (product promise), development guidelines and 1. authoritative source mandate.
mq commit.instructions.md
Instructions for harehare/mq: Use the following format for commit messages.