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/benramins/otto/claude-mdgit clone --depth 1 https://github.com/BenRamins/ottoWrote 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/benramins/otto/claude-md)<a href="https://agentmods.dev/instructions/benramins/otto/claude-md"><img src="https://agentmods.dev/badge/instructions/benramins/otto/claude-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 | $0.01162 | $0.01162 |
| Opus 5 | $0.00581 | $0.00581 |
| Sonnet 5 | $0.00232 | $0.00232 |
| Haiku 4.5 | $0.00116 | $0.00116 |
Grade A, and why
otto 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 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.
How it starts
The opening of the file, as written. The whole thing — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md — Agent Operating Rules
How agents work in this repo (otto-dev, the framework itself) — not the rules Otto installs into a target repo. Auto-loaded every session. Keep it tight; promote durable
learnings/entries up here via/otto-distill.
What this repo is
Otto is a routing and orchestration layer, not an implementation. It orchestrates three
read-only sources — SK spec-kit · AS agent-skills · PK pocock-skills — and ships the
install recipe, never the implementations. See sources/sources.md.
There is no build, no runtime, no dependency tree. The artifacts are Markdown, one YAML registry,
and shell/PowerShell scripts. Layers: sources (0) → registry/capabilities.yaml (1)
→ router/ (2) → workflows/spine.md (3) → recipes (4).
Commands
- Self-consistency:
bash install/verify-otto.sh— audits Otto against itself (invocation drift, registered-skill drift, stamped-pin drift, dashboard↔registry drift). Run after any registry, dashboard, or skill-unit edit. - Audit one run:
bash install/otto-gate.sh specs/NNN— checks a run's gate trail is auditable (rows exist, PASS/WAIVED name checkable evidence, no stage stuck on FAIL). - Register skill units:
install/register-skills.{ps1,sh}— re-run after editing any Otto skill unit, or the host keeps serving the old copy (known drift lesson;verify-otto.shcheck 7 exists because of it). - Per-repo setup:
install/otto-init.{ps1,sh}— scaffolds a target repo end-to-end. - Hooks:
otto-session-start.{ps1,sh},otto-learnings-nudge.{ps1,sh}.
Scripts come in .ps1 + .sh mirrors — edit both or neither. otto-gate.sh and
verify-otto.sh are bash-only today.
Conventions
registry/capabilities.yamlis canonical. DESIGN.md §5 is the human-readable view of it — when they disagree, the YAML wins and DESIGN is the bug.- Invocation rendering: the registry writes spec-kit's dotted form (
/speckit.plan); Claude Code renders it hyphenated (/speckit-plan). Emit the hyphenated form to the user; keep the dotted form in the registry. - Windows encoding (load-bearing): ship ASCII-only
.ps1; write every file UTF-8 without BOM. A BOM breaks YAML-frontmatter parsing of anySKILL.md-like file. - Assessments go in
assessments/YYYY-MM-DD-<slug>.mdwith the house shape: meta blockquote → Verdict → evidence tables → Coverage (what was not scanned, explicitly) → Recommendation. - Deferred work goes in
BACKLOG.mdwith an act-on trigger — an observable event, not a date. A row without a trigger is a wish. learnings/global.mdentries are hints-to-verify, not facts. Re-check any named file, flag, or version before acting on one.
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 · 76 lines · 1,162 tokens per session scan A c3cfaec90a67
otto CLAUDE.md is an instructions file published in the GitHub repository BenRamins/otto (12 stars, last pushed 21d ago), licensed MIT. It adds 1,162 tokens to every session, about $0.0058 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
workflow AGENTS.md
AGENTS.md instructions for vercel/workflow, covering agent instructions, architecture, core components, workflow execution model and development commands.
a-i--skills 03-development-workflow.instructions.md
Instructions for organvm-iv-taxis/a-i--skills, covering development workflow instructions, no repo-wide build system, common development commands, after adding/modifying skills and validation before committing.
a-i--skills GEMINI.md
Instructions for organvm-iv-taxis/a-i--skills, covering gemini.md — a-i--skills, system context (auto-generated — do not edit), edges, siblings in orchestration and governance.
team-shinchan AGENTS.md
Instructions for seokan-jeong/team-shinchan, covering agents.md - team-shinchan agent map, layer architecture, agent registry, call flow and quick reference.
simple-agent-manager AGENTS.md
AGENTS.md instructions for raphaeltm/simple-agent-manager, covering sam — agent supplement, agent configuration cross-reference, skills, review / specialist skills and reference skills.
nika AGENTS.md
AGENTS.md instructions for supernovae-st/nika, covering agents — nika diamond, repository snapshot, what to read first, hard rules (non-negotiable) and crate admission: 12 gates.