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/ghillb/swarmux/agents-mdgit clone --depth 1 https://github.com/ghillb/swarmuxWrote 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/ghillb/swarmux/agents-md)<a href="https://agentmods.dev/instructions/ghillb/swarmux/agents-md"><img src="https://agentmods.dev/badge/instructions/ghillb/swarmux/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 | $0.00346 | $0.00346 |
| Opus 5 | $0.00173 | $0.00173 |
| Sonnet 5 | $0.00069 | $0.00069 |
| Haiku 4.5 | $0.00035 | $0.00035 |
Grade A, and why
swarmux 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 4d 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.
Copies of this mod
1 near-identical copy found in the catalogue:
- swarmux AGENTS.md — 100% identical, 0 lines differ
What it actually says
Agent Instructions
All CLI implementation and CLI UX decisions in this repo must follow these rules:
- JSON-first contract
- Every command must support machine-readable output.
- Prefer JSON as the default for agent-facing paths.
- Human-friendly text output is secondary and must not change machine semantics.
- Deterministic interfaces
- Keep flags, field names, and output shapes stable.
- Avoid ambiguous parsing and context-dependent behavior.
- Treat CLI output as an API contract.
- Non-interactive by default
- No prompts for normal execution paths.
- No hidden defaults that require human interpretation.
- Everything required for execution must be explicit in flags or JSON payloads.
- Separation of data and presentation
- Do not encode presentation-only strings into machine fields.
- Keep structured data in structured fields; render formatting only in text mode.
- Script/agent safety
- Commands should be composable, predictable, and idempotent where possible.
- Provide clear exit codes and actionable errors.
- Prefer explicit validation errors over silent coercion.
- Verb-based task model
- Use clear, action-oriented subcommands (
submit,start,show, etc.). - Keep command behavior narrow and single-purpose.
- Validation and simulation
- Validate inputs defensively (agents hallucinate malformed values).
- Support dry-run/preview for mutating operations whenever feasible.
- Runtime semantics
tuiis a first-class runtime besideheadlessandmirrored.- Runtime controls task/session execution only. Popup/window presentation belongs in tmux bindings or operator workflow, not in the core CLI contract.
If there is a conflict with existing CLI patterns, align with this rule set.
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.
- 4d ago First seen · 42 lines · 346 tokens per session scan A 0390c4cc9a5d
swarmux AGENTS.md is an instructions file published in the GitHub repository ghillb/swarmux (5 stars, last pushed 5mo ago), licensed MIT. It adds 346 tokens to every session, about $0.0017 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
CLITrigger CLAUDE.md
Claude Code instructions for HyperAITeam/CLITrigger, covering claude.md, project overview, commands, development (runs server + client concurrently) and build.
ensemblr AGENTS.md
Instructions for ensemblr-hq/ensemblr, covering agent instructions, project naming, app scaffolding requires current official docs, package manager policy and biome policy.
wiff AGENTS.md
Instructions for xxxoooxoxo/wiff, covering agents.md, orchestrating with wiff (driving it from a harness) and working on this codebase.
multi-agent-loop-kit AGENTS.md
Instructions for anshulixyz/multi-agent-loop-kit, covering agents.md — multi-agent ownership registry, two meanings of “agent”, core rules, project lead / operator and beacon — loop intelligence agent.
agent-semaphore CLAUDE.md
Claude Code instructions for alwh1te/agent-semaphore, covering agent-semaphore — repository rules, hard rules, stack and commands and conventions.
forge AGENTS.md
Instructions for automagik-dev/forge, covering genie agent framework, core identity, core purpose, task context (auto-loaded) and product documentation.