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/isaiascope/ai/agents-mdgit clone --depth 1 https://github.com/IsaiaScope/aiWrote 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/isaiascope/ai/agents-md)<a href="https://agentmods.dev/instructions/isaiascope/ai/agents-md"><img src="https://agentmods.dev/badge/instructions/isaiascope/ai/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 | $0.02398 | $0.02398 |
| Opus 5 | $0.01199 | $0.01199 |
| Sonnet 5 | $0.00480 | $0.00480 |
| Haiku 4.5 | $0.00240 | $0.00240 |
Grade B, and why
ai AGENTS.md scanned grade B with 1 finding 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
Copies `config/CLAUDE.md` → `~/CLAUDE.md` and `config/AGENTS.md` → `~/.codex/AGENTS.md`, installs upstream skill packs via `npx skills@latest`, symlinks the local `IsaiaScope/ai` skills directly into both supported agent How it starts
The opening of the file, as written. The whole thing — 127 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
This file provides guidance to coding agents (Claude Code, Codex) working with code
in this repository. CLAUDE.md is a symlink to this file — edit here.
Install / Update
node scripts/install.js
Copies config/CLAUDE.md → ~/CLAUDE.md and config/AGENTS.md → ~/.codex/AGENTS.md, installs upstream skill packs via npx skills@latest, symlinks the local IsaiaScope/ai skills directly into both supported agents' skills dirs (Claude-side → ~/.claude/skills/, Codex-side → ~/.codex/skills/), and owns the tracker's two hooks in ~/.claude/settings.json (see scripts/agent-hooks.js). No build step, no package.json. The whole suite is one command, and it lives in docs/iso/config.json under test.command — that is the command /iso-review's phase gate runs, so running anything else locally checks less than CI does. It covers node --test scripts/*.test.js, every *.test.sh under scripts/ and skills/*/scripts/{,lib/,adapters/}, and every skills/*/scripts/test_*.py.
Architecture
config/
AGENTS.md — global agent instructions (copied to ~/.codex/AGENTS.md on install)
CLAUDE.md — symlink to AGENTS.md (copied to ~/CLAUDE.md on install)
skills/ — prefix routes each skill to a marketplace plugin (iso-* → eng, social-* → social)
iso-config/SKILL.md — the Iso config every iso-* skill reads
iso-config/scripts/lib/branch.sh — shared branch vocabulary: protected test, name derivation, the gate
iso-config/scripts/lib/track.sh — the one seam to the tracker: iso_track, iso_track_path; never fatal
iso-issue-tracking/SKILL.md — work tracker, reached through a swappable adapter
iso-ai-init/SKILL.md — initialize a repo with IsaiaScope AI defaults
iso-init-repo/SKILL.md — initialize repo governance (branches, CI, hooks)
iso-plan/SKILL.md — planning pipeline orchestrator
iso-write/SKILL.md — TDD plan executor on a feature branch, no commits
iso-review/SKILL.md — three phases over the branch: architecture, simplify, review
iso-spawn/SKILL.md — spawn a codex/claude agent in a herdr tab
iso-readme/SKILL.md — write/refine READMEs in house style, commit + push
social-notebooklm/SKILL.md — research-first NotebookLM notebook + scaletta for a video
social-new-video/SKILL.md — one-shot launcher: spawns a tab that runs social-notebooklm
scripts/
install.js — deploys config files + installs skill packs globally
agent-hooks.js — writes the tracker hooks into ~/.claude/settings.json (list: skills/iso-issue-tracking/scripts/hooks.json)
skills-manifest.js — filesystem → skill catalog: discovery, prefix routing (PLUGINS), manifest projection
dispatch-integrity.test.sh — every verb a script dispatches must resolve to a defined function
portability.test.sh — no skill may carry the machine it was written on (paths, package manager, agent config dir)
skill-refs.test.sh — every skill name referenced in live prose has a directory (survives a rename's blind spots)
.claude-plugin/
marketplace.json — Claude marketplace catalog (2 plugins → ./plugins/<name>)
.agents/plugins/
marketplace.json — Codex marketplace catalog (2 plugins → ./plugins/<name>)
plugins/ — one subdir per marketplace plugin (routed by skill prefix)
isaiascope-eng/ — engineering plugin (iso-* skills)
.claude-plugin/plugin.json — Claude manifest (skills array, regenerated on install)
.codex-plugin/plugin.json — Codex manifest (skills: "./skills/", whole-dir)
skills → ../../../skills/iso-* — per-skill symlinks, regenerated on install
isaiascope-social/ — social plugin (social-* skills)
.claude-plugin/plugin.json — Claude manifest (skills array, regenerated on install)
.codex-plugin/plugin.json — Codex manifest (skills: "./skills/", whole-dir)
skills → ../../../skills/social-* — per-skill symlinks, regenerated on install
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 · 127 lines · 2,398 tokens per session scan B 2b3262e71ece
ai AGENTS.md is an instructions file published in the GitHub repository IsaiaScope/ai (2 stars, last pushed 7d ago), licensed MIT. It adds 2,398 tokens to every session, about $0.0120 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
claude-skill-registry CLAUDE.md
Claude Code instructions for majiayu000/claude-skill-registry, covering claude.md, repo map, how to decide where to edit, generated vs owned files in main and publish flow.
pivoshenko.ai CLAUDE.md
Instructions for pivoshenko/pivoshenko.ai, covering claude.md, what this repo is, commands, asset layout and the sync contract and frontmatter contract.
skern AGENTS.md
Instructions for devrimcavusoglu/skern, covering agents.md — skern development guide, project overview, scope, package responsibilities and design decisions.
skern CLAUDE.md
Instructions for devrimcavusoglu/skern: All agent instructions live in AGENTS.md. Refer to that file for project overview, build commands, conventions, architecture, and workflow details.
skills AGENTS.md
Instructions for arcjet/skills, covering agents.md, subagent prompts for test runs, aggregation gotchas, capturing subagent transcripts and eval viewer.
skills AGENTS.md
AGENTS.md instructions for payloadcms/skills, covering payload skills repository, repository overview, creating a new skill, directory structure and naming conventions.