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/bitmia-ai/redeye/claude-mdgit clone --depth 1 https://github.com/Bitmia-ai/RedEyeWrote 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/bitmia-ai/redeye/claude-md)<a href="https://agentmods.dev/instructions/bitmia-ai/redeye/claude-md"><img src="https://agentmods.dev/badge/instructions/bitmia-ai/redeye/claude-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 | $0.01487 | $0.01487 |
| Opus 5 | $0.00744 | $0.00744 |
| Sonnet 5 | $0.00297 | $0.00297 |
| Haiku 4.5 | $0.00149 | $0.00149 |
Grade A, and why
RedEye 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 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 — 98 lines — stays where its author put it; the contents beside it link to each section on GitHub.
RedEye
A Claude Code plugin that continues your development work overnight. Leave a task list, walk away, come back to shipped features.
Commands
| Command | Description |
|---|---|
/redeye:init |
Scaffold .redeye/ control files in your project |
/redeye:start |
Start (or resume) the autonomous dev loop |
/redeye:stop |
Gracefully stop after the current phase |
/redeye:status |
What's done, what's next, open questions |
/redeye:tasks |
View or add tasks |
/redeye:steer |
Add a directive ("focus on tests", "skip mobile") |
/redeye:brainstorm |
Think through an idea, turn it into a task |
/redeye:log |
Recent iteration history |
/redeye:schedules |
Manage recurring tasks |
/redeye:pause |
Pause after the current task cycle |
How It Works
RedEye pre-computes a .redeye/digest.json from all control files (zero LLM tokens — pure bash+jq). The CTO agent reads this digest each iteration, decides which phase to run, and dispatches an isolated sub-agent. One phase per iteration, sequentially.
Main cycle: TRIAGE → PLAN → BUILD → REVIEW → DEPLOY → VERIFY → MERGE → next item
Side phases (triggered by conditions in the digest):
- STABILIZE — environment broken, focuses on restoring health (max 3 attempts)
- INCORPORATE — user answered questions in
.redeye/inbox.md, adjusts existing work - SCHEDULES — recurring tasks overdue
On empty task list: RedEye stops and waits for new tasks.
On review failure: Routes back to BUILD. After 3 failed cycles, parks the task and moves on.
On needing user input: Writes question to .redeye/inbox.md with a sensible default. Either proceeds with the default or blocks the task and picks up the next one. Never hallucinates an answer.
Control Files
All live in .redeye/ in the user's project. /redeye:init adds every entry below to the project's .gitignore — the control files are runtime state and are never committed.
| File | Purpose |
|---|---|
config.md |
Project config — name, stack, deploy commands, vision |
tasks.md |
Task list, worked top-down |
steering.md |
Tactical directives |
inbox.md |
Questions from RedEye, answers from the user |
status.md |
Current progress |
changelog.md |
What shipped |
feedback.md |
RedEye's retros on its own work |
schedules.md |
Recurring tasks |
state.json |
Runtime state |
digest.json |
Pre-computed digest (regenerated each iteration) |
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 · 98 lines · 1,487 tokens per session scan A eff3df1d0d68
RedEye CLAUDE.md is an instructions file published in the GitHub repository Bitmia-ai/RedEye (6 stars, last pushed 2mo ago), licensed MIT. It adds 1,487 tokens to every session, about $0.0074 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 2026-08-31.
Other instructions, from other repositories
aiwg CLAUDE.md
Claude Code instructions for jmagly/aiwg, covering provider workspace bootstrap, provider workspace bootstrap and aiwg artifact output policy.
aiwg AGENTS.md
AGENTS.md instructions for jmagly/aiwg, covering agents.md, provider workspace bootstrap and framework context.
nightshift AGENTS.md
AGENTS.md instructions for orwa-mahmoud/nightshift: AI-assisted contributions are welcome. The same correctness, security, licensing, testing, and quality standards apply regardless of which tools are used.
erpaval CLAUDE.md
Instructions for theagenticguy/erpaval, covering erpaval-plugin, repo structure, version management, tools and hooks.
claude-sdlc CLAUDE.md
Claude Code instructions for lantisprime/claude-sdlc, covering claude.md, what this repo is, design intent — read this before "improving" anything, eat your own dog food and hook strictness philosophy.
yoink AGENTS.md
Instructions for theogbrand/yoink, a project described as: yoink is an AI agent that removes complex dependencies by reimplementing only what you need.