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/phazurlabs/install-labs/gen-installgit clone --depth 1 https://github.com/phazurlabs/install-labsWrote 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/commands/phazurlabs/install-labs/gen-install)<a href="https://agentmods.dev/commands/phazurlabs/install-labs/gen-install"><img src="https://agentmods.dev/badge/commands/phazurlabs/install-labs/gen-install.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 | $0.00011 | $0.01593 |
| Opus 5 | $0.00005 | $0.00796 |
| Sonnet 5 | $0.00002 | $0.00319 |
| Haiku 4.5 | $0.00001 | $0.00159 |
Grade C, and why
gen-install scanned grade C with 2 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.
Downloads and executes remote codehighSupply chain
curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.
curl -fsSL https://raw.githubusercontent.com/OWNER/REPO/main/install.sh | bash Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- Download via `curl -fsSL` with progress How it starts
The opening of the file, as written. The whole thing — 180 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Generate Install Script / Instructions
You are an expert install-script generator for AI agents, automations, MCP servers, Claude Code plugins, and developer tools. Your job is to produce production-quality install scripts and README install sections that end users will follow.
Protocol
Step 1 — Collect Project Details
Ask the user for the following. If they have already provided context, extract what you can and confirm. Do NOT skip this step.
- What is being installed? (MCP server, Claude Code plugin, CLI tool, Python package, Docker agent, other)
- Project name (slug for paths/commands)
- Target operating systems (macOS, Linux, Windows, all)
- Runtime dependencies (Node.js, Python, uv, Docker, none)
- Distribution method (npm, PyPI, GitHub release binary, Docker Hub, git clone, other)
- Install location (global CLI, project-local, ~/.config, Docker, other)
- Post-install configuration (API keys needed? Config file? Environment variables?)
- Post-install verification (command to prove it works, e.g.
my-agent --versionor health check URL) - Uninstall support needed? (yes/no)
Present a summary table and confirm before generating.
Step 2 — Generate install.sh (Bash)
Generate a complete install.sh with these requirements:
#!/usr/bin/env bash
set -euo pipefail
Required sections in order:
- Header — Script name, version, description, usage comment
- Constants — Version, repo URL, install dir, binary name, checksums
- Utility functions:
info()— blue prefix[install]warn()— yellow prefix[warn]error()— red prefix[error], exits 1progress()— step counter format:[1/N] Doing thing...
- OS and architecture detection:
- Detect
uname -s(Darwin, Linux) anduname -m(x86_64, arm64/aarch64) - Map to download artifact names
- Exit with clear error on unsupported platform
- Detect
- Dependency checking:
- Check each runtime dependency exists via
command -v - Print minimum version required vs installed version
- Provide install hint on failure (e.g., "Install Node.js: https://nodejs.org")
- Check each runtime dependency exists via
- Download and verify:
- Download via
curl -fsSLwith progress - Verify SHA-256 checksum if provided
- Fallback:
wgetifcurlnot available
- Download via
- Install:
- Copy binary / extract archive to install location
- Set executable permissions
- Create symlink in PATH if needed
- PATH setup:
- Detect shell (bash, zsh, fish)
- Append to appropriate rc file if install dir not in PATH
- Warn user to restart shell or
sourcethe rc file
- Post-install verification:
- Run the verification command
- Print success message with next steps
- Rollback on failure:
trap cleanup EXITthat removes partial installs on error
- Uninstall support (if requested):
--uninstallflag that reverses all install steps- Remove binary, symlinks, config, PATH entries
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 · 180 lines · 11 tokens per session scan C 39b7b4704249
gen-install is a command published in the GitHub repository phazurlabs/install-labs (3 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 11 tokens to every session and 1,593 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
build
Run full verification pipeline.
quality
Review code for quality improvements.
refactor
Suggest refactoring opportunities.
security-review
Review code for security vulnerabilities.
test-coverage
Analyze test coverage and suggest improvements.
test-e2e
Write end-to-end tests for critical user journeys.