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/hexxt-git/umm/claude-mdgit clone --depth 1 https://github.com/hexxt-git/ummWhat 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 | $0.02582 | $0.02582 |
| Opus 5 | $0.01291 | $0.01291 |
| Sonnet 5 | $0.00516 | $0.00516 |
| Haiku 4.5 | $0.00258 | $0.00258 |
Grade A, and why
umm CLAUDE.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 — 165 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
What umm is
A CLI that answers a question in a compact, terminal-shaped way instead of conversational prose. It does almost nothing itself: it hands the question plus a skill (a big prompt) to whichever coding agent the user configured, then renders the markdown that comes back. The product is the skill, not the code.
The two markdown files are the real source of truth
SKILL.md(repo root) — the portable "brain". Defines the entire output shape and behavior. It is validated prompt content: reword it deliberately, not casually, and re-test against real queries after changing it. Root placement is intentional — it makes the repo installable as a standalone agent skill (npx skills add <owner>/umm).cli.md— a CLI-only addendum bolted on at prompt-build time. It tells the agent it has no edit permissions and to redirect the user (usually: run the agent directly instead of through umm). It is deliberately not part of the standalone skill, so anpx skillsinstall never sees it.
Both are excluded from Prettier (.prettierignore) because their exact
wording/whitespace is functional.
Build model: the skill is embedded, not read at runtime
scripts/embed-skill.ts reads SKILL.md + cli.md, strips frontmatter, and
writes src/skill.generated.ts (exports SKILL, CLI_ADDENDUM).
src/skill.generated.tsis generated and gitignored. A fresh clone must runnpm run build:skill(orbuild/typecheck, which do it) before the CLI will run.- Editing
SKILL.md/cli.mdrequires a rebuild to take effect.
Imports use .js, but the source is .ts
Relative imports are written with .js specifiers (./agents.js) even though
the files are .ts. This is the tsc ESM convention (tsc resolves .js → .ts
source and emits real .js). Consequence:
node src/index.tsdoes NOT work (Node won't rewrite.js→.ts).- Dev runs through Bun:
bun src/index.ts <question>(ornpm run dev). - npm ships the compiled
dist-npm/, not the source.
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 · 165 lines · 2,582 tokens per session scan A 85a305c29567
umm CLAUDE.md is an instructions file published in the GitHub repository hexxt-git/umm (22 stars, last pushed 26d ago), licensed MIT. It adds 2,582 tokens to every session, about $0.0129 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 AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
buildNext
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
spec-kit AGENTS.md
Instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
langchain AGENTS.md
Instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.