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/tmchow/illo-skill/agents-mdgit clone --depth 1 https://github.com/tmchow/illo-skillWhat 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 | $0.04931 | $0.04931 |
| Opus 5 | $0.02465 | $0.02465 |
| Sonnet 5 | $0.00986 | $0.00986 |
| Haiku 4.5 | $0.00493 | $0.00493 |
Grade A, and why
illo-skill 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 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 — 368 lines — stays where its author put it; the contents beside it link to each section on GitHub.
illo-skill
The canonical home of illo, a cross-platform AI agent skill (SKILL.md
format) installable through runtime-specific lanes: the generic skills CLI
(npx skills add tmchow/illo-skill --skill illo, works across Claude
Code/Cursor/Codex and other Agent-Skills runtimes), Hermes (GitHub directory
identifier: hermes skills install tmchow/illo-skill/illo or
/skills install tmchow/illo-skill/illo; raw SKILL.md URLs are valid only
for single-file fallback installs — never for illo, which is multi-file), and
OpenClaw (ClawHub via openclaw skills install illo). Document lanes side by
side; never present one as the only path.
This guide is for anyone (human or agent) editing the repo. Keep it accurate when conventions change.
illo was extracted from tmchow/agent-skills
(June 2026, full history preserved). That repo keeps a frozen copy and must
keep serving its _assets/illo/ raw URLs — they are baked into
already-installed copies — but all development and publishing happen here.
Repo layout
The skill lives in skills/illo/, deliberately not at the repo root.
The top-level skills/ folder matches the layout convention of the
canonical skill repos (anthropics/skills, openai/skills,
vercel-labs/agent-skills) and keeps a future Claude Code plugin lane
possible (.claude-plugin/ + skills/). Installers copy the entire skill
directory verbatim, so the skill dir must contain only what every install
should ship; promoting SKILL.md to the root would make _assets/ and
repo meta ship with every install and blow the size budget. Note the
install identifiers do not carry the skills/ segment — skills.sh
indexes by skill name (tmchow/illo-skill/illo), the same way
openai/skills/skill-creator hides its skills/ folder.
skills/illo/SKILL.md— required. The agent-facing instructions.skills/illo/README.md— required. The human-facing landing page (below).skills/illo/references/— deep material loaded on demand, including the seventeen look definitions inreferences/styles/.skills/illo/scripts/— the engine (illo.py) and the Hermes asset-repair preflight.skills/illo/assets/— bundled binary assets pluschecksums.txt, a generated manifest (never edit by hand; see Binary assets below)._assets/illo/— docs-only images linked by raw URL (calibration examples, README embeds). They live outside the skill directory so they never ship with installs..github/— the ClawHub publish and asset-checksum workflows and their scripts.- Root
README.md— the repo landing page; rootLICENSE— MIT.
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 · 368 lines · 4,931 tokens per session scan A 16799c8f6151
illo-skill AGENTS.md is an instructions file published in the GitHub repository tmchow/illo-skill (369 stars, last pushed 4d ago), licensed MIT. It adds 4,931 tokens to every session, about $0.0247 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 2026-08-30.
Other instructions, from other repositories
codex-image-context-runtime AGENTS.md
Instructions for shixinnt/codex-image-context-runtime, covering agents.md, public boundary, runtime contract and changes.
openai-compatible-imagegen AGENTS.md
Instructions for Syh1906/openai-compatible-imagegen, covering agent instructions, project scope, package management, checks and module boundaries.
diffusers AGENTS.md
AGENTS.md instructions for huggingface/diffusers, a project described as: 🤗 Diffusers: State-of-the-art diffusion models for image, video, and audio generation in PyTorch.
comfyui_mcp AGENTS.md
Instructions for hybridindie/comfyui_mcp, covering agents.md, independent & harness-agnostic, project overview, tech stack and project structure.
OGAD AGENTS.md
AGENTS.md instructions for off-grid-ai/OGAD, covering off grid ai desktop — agent guide, design — desktop-first, off grid brand, use the screen real estate — desktop density rules, what this app is and stack.
atet AGENTS.md
Instructions for hraness/atet, a project described as: Local-first AI media generation and video editing toolkit for coding agents, with a Bun CLI, TypeScript SDK, MCP server, and Agent Skill.