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/bitmia-ai/redeye/startgit clone --depth 1 https://github.com/Bitmia-ai/RedEyeWhat 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 | $0.00007 | $0.01109 |
| Opus 5 | $0.00003 | $0.00554 |
| Sonnet 5 | $0.00001 | $0.00222 |
| Haiku 4.5 | $0.00001 | $0.00111 |
Grade A, and why
start 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 — 105 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Start RedEye
Validate the project is ready and start the Ralph Loop.
Step 1: Validate Setup
Check that these files exist in the project root:
.redeye/config.md— if missing, tell user to run/redeye:initfirst.redeye/config.mdmust contain a## Visionsection with content (not just the placeholder). If missing or empty, tell user to add a vision..redeye/tasks.md— if no items, tell user to add at least one item.redeye/state.json— if missing, tell user to run/redeye:init
If any validation fails, stop and guide the user.
Step 2: Check Resume State
Read .redeye/state.json:
- If
iteration> 0: this is a resume- Show: "Resuming from iteration {n} ({phase})"
- Show last 3 entries from
iteration_log - Show health/confidence
- If
iteration= 0: this is a fresh start- Show: "Starting fresh"
Step 3: Show Summary
Present to the user:
- Vision: first 100 chars of ## Vision section in .redeye/config.md
- Tasks: count of items (CEO Requests + Discovered + Triaged)
- Health: confidence level, env status
- Next: what the team will work on first (highest-priority planned item, or first CEO Request if nothing triaged)
Step 4: Start Loop
CRITICAL: You MUST run the start-loop.sh script. Do NOT skip this. It creates the .claude/ralph-loop.local.md file that enables autonomous looping. Without it, the session will stop after each response instead of continuing automatically.
"${CLAUDE_PLUGIN_ROOT}/scripts/start-loop.sh" .
If CLAUDE_PLUGIN_ROOT is not set, find the plugin root by looking for .claude-plugin/plugin.json in parent directories or common plugin locations.
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 · 105 lines · 7 tokens per session scan A d9e95a77dc76
start is a command published in the GitHub repository Bitmia-ai/RedEye (6 stars, last pushed 2mo ago), licensed MIT. It adds 7 tokens to every session and 1,109 once invoked, about $0.0000 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 commands, from other repositories
squad
Spawn a self-coordinating squad to design and coordinate a sub-team.
workflow-claude-commands
Track Claude Code commands report changes and find what needs updating.
agent-farm
The afx (agent-farm) command manages multi-agent orchestration for software development. It spawns and manages builders in isolated git worktrees.
tool-connector
Connect any tool you use at work — internal company tools, custom-built systems, and commercial tools like Slack, Confluence, Jira, Linear, GitHub, Outlook, Datadog, and PagerDuty.
conventions-check
Check AGENTS.md, CLAUDE.md, and the projected agent artifacts (skills, subagents, commands) against the repo's conventions.
parallel-task
Executes a structured plan in parallel waves, spawning appropriate agents for each task based on dependencies and confidence routing.