agentmods.dev

onboard CLAUDE.md

onboard CLAUDE.md is an instructions file for coding agents from sdsrss/onboard. It costs 4,420 tokens per session, scanned E, original, MIT.

Repository instructions for the onboard Claude Code skill, a reusable workflow that helps agents inspect and prepare a project. They document its phases, versioning, checks, sharing modes, and plugin recommendations.

In plain words
What is it for?
Use them when developing, versioning, testing, or documenting the onboard skill and its installation or removal workflow.
Why use it?
They explain how this skill is structured and validated, so an agent can change it without confusing the skill itself with a normal application.

Instructions file

Installs and runs on its own, but its text points at files inside the plugin that ships it — anything it tells you to read at a ${CLAUDE_PLUGIN_ROOT} path is only there once the plugin is installed.

not rated 2 3mo ago E scan 4,420 tokens original MIT
View the source file sdsrss/onboard
Install

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.

agentmods
npx agentmods add instructions/sdsrss/onboard/claude-md
Clone the repo
git clone --depth 1 https://github.com/sdsrss/onboard

Wrote 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.

agentmods badge for onboard CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/sdsrss/onboard/claude-md.svg)](https://agentmods.dev/instructions/sdsrss/onboard/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/sdsrss/onboard/claude-md"><img src="https://agentmods.dev/badge/instructions/sdsrss/onboard/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 4,420 This file is loaded in full into every session.
When invoked 4,420 The same file — it is already loaded in full.
Security scan E 3 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.04420 $0.04420
Opus 5 $0.02210 $0.02210
Sonnet 5 $0.00884 $0.00884
Haiku 4.5 $0.00442 $0.00442

Measured 4d ago against content hash 087f2879f252, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade E, and why

onboard CLAUDE.md scanned grade E with 3 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.

| `guard-bash.sh` | PreToolUse / `Bash` | Deny dangerous shell patterns (`rm -rf /`, force-push to main, `curl \| sh`, …) |

Recursive force deletehighDestructive command

rm -rf with a variable or a broad path is one typo away from removing the wrong tree.

| `guard-bash.sh` | PreToolUse / `Bash` | Deny dangerous shell patterns (`rm -rf /`, force-push to main, `curl \| sh`, …) |

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

| `guard-bash.sh` | PreToolUse / `Bash` | Deny dangerous shell patterns (`rm -rf /`, force-push to main, `curl \| sh`, …) |
CLAUDE.md · 120 lines

How it starts

The opening of the file, as written. The whole thing — 120 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

What this repo is

This repo is not an application — it is the /onboard Claude Code Skill itself (current version: v3.2.0). The "code" being maintained is:

  • skills/onboard/SKILL.md (v2.10 moved here from repo root) — the workflow spec defining a 10-phase onboarding protocol (Phase 0 / 0.5 / 1 / 1.5 / 1.7 / 2 / 2.5 / 3-8), Doctor Mode (--doctor, v2.5+; D1-D15 checks), Mode model (--local-only default vs --share, v2.6+), Install Plan / Plugin recommendation matrix (v2.7+), Uninstall Mode + marker/snapshot protocols (v2.8+), and Plugin marketplace install-source detection (v2.9+). v3.0 起 Phase 7 / Uninstall Mode / 状态文件结构 三段抽到 sub-file；v3.2 起新增 Plugin recommendation matrix 第 4 段 (skills/onboard/phases/phase-7.md / skills/onboard/phases/uninstall.md / skills/onboard/references/state-schema.md / skills/onboard/references/recommendations.md)；SKILL.md 原位置保留 sentinel header + 4-8 bullet summary + Read sub-file.md 指令；consumer Claude 进入对应 phase 前必须 Read sub-file（元规则 27）。This file's frontmatter (name: onboard, disable-model-invocation: true, allowed-tools: Read, Glob, Grep) registers it as a Skill — at skills/<name>/ for plugin discovery, or under standalone .claude/skills/onboard/ after install.sh.
  • skills/onboard/hooks/*.sh (v2.10 moved) — 4 hook scripts; git index mode 100755 (executable).
  • skills/onboard/settings.template.json + skills/onboard/settings.local.template.json (v2.10 moved) — reference configs for --share / --local-only modes; co-located with SKILL.md.
  • install.sh (v2.8, v2.10 redesigned) — universal installer (install/update/uninstall/doctor); uses staging cache at ~/.claude/.cache/onboard-source/ for fast updates; copies skills/onboard/* into install target.
  • .claude-plugin/plugin.json (v2.8, v2.9 schema-corrected) — Claude Code plugin manifest, canonical schema only.
  • .claude-plugin/marketplace.json (v2.9) — Claude Code plugin marketplace catalog at repo root; required for /plugin marketplace add sdsrss/onboard to find the plugin.
  • tests/run.sh + tests/integration/*.sh — in-repo sandbox tests for plugin install / hook execution / installer round-trip / SKILL.md link consistency / hook runtime behavior / v2.11 uninstall modes / v2.12 state-schema validation (224 assertions across 10 integration tests: plugin-install.sh 24 + hook-mirror.sh 33 + install-roundtrip.sh 38 + skillmd-links.sh 21 + hook-behavior.sh 55 + cc-plugin-detection.sh 8 + prepare-script-detection.sh 2 + sync-versions-detection.sh 1 + uninstall-modes.sh 21 + state-schema.sh 21); run via bash tests/run.sh. v2.11.1 adds scripts/verify-counts.sh to auto-check the headline+per-test claims here / in CHANGELOG against actual pass: counts — drift triggers non-zero exit. v2.11.2 adds scripts/release-preflight.sh which bundles verify-counts.sh + version-bump completeness + git update-index --chmod=+x on new .sh + staged-file sanity into a single pre-commit check. v2.11.3 adds .env boundary-anchored deny in guard-bash.sh (H1; 12 new hook-behavior assertions) + mirror-hooks.sh cmp -s fast path (L1; 4 new hook-mirror assertions) + inline _keeper_command_after_skill_uninstall doc fields in settings.local.template.json (M5) + install.sh Iron Law 14 stage-cache exemption comment (M4) + README "never auto-installs" 设计承诺 section. v2.12.0 adds ONBOARD_FORBIDDEN_COMMANDS newline-separated regex env in guard-bash.sh (H2; 11 new hook-behavior assertions) + per-stack lint_timeout_sec / typecheck_timeout_sec / format_timeout_sec / format_check_timeout_sec in stacks.json (H3; 4 new hook-behavior assertions) + scripts/validate-state.sh runtime state-schema validator + state-schema.sh integration test (C2; 21 new assertions). Persisted post-v2.10 from the prior /tmp/onboard-plugin-test.sh; hook-behavior.sh added in v2.10.2; cc-plugin-detection.sh / prepare-script-detection.sh / sync-versions-detection.sh / uninstall-modes.sh added in v2.11.0 for the 8 P-A items; state-schema.sh added in v2.12.0 for C2.
  • README.md — install + usage guide for end users.
  • CHANGELOG.md — version history.

Read the full file on GitHub · 120 lines

Changes

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.

  1. 4d ago First seen · 120 lines · 4,420 tokens per session scan E 087f2879f252

Subscribe to this mod's changes

onboard CLAUDE.md is an instructions file published in the GitHub repository sdsrss/onboard (2 stars, last pushed 3mo ago), licensed MIT. It adds 4,420 tokens to every session, about $0.0221 per session on Opus 5. A static security scan graded it E with 3 findings (downloads and executes remote code, recursive force delete, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

ai-toolingautomationci-alignmentclaude-codeclaude-code-pluginclaude-code-skill
Related

Other instructions, from other repositories

recursive-decomposition-skill AGENTS.md

Instructions for massimodeluisa/recursive-decomposition-skill, covering agents: recursive-decomposition skill, read these first (mandatory), language policy, non-negotiables and commands.

massimodeluisa/recursive-decomposition-skill · 321 tokens

stockbit-mcp CLAUDE.md

Claude Code instructions for INo-xious/stockbit-mcp, covering claude.md, what this is, commands, the map and three invariants. do not break them; each has a test.

INo-xious/stockbit-mcp · 1,337 tokens

devin-handoff AGENTS.md

Instructions for club-cog/devin-handoff, covering devin handoff — agent guide, what this is, when to hand off, finding the script and how to use.

club-cog/devin-handoff · 542 tokens

plumb-line AGENTS.md

Instructions for slopstopper/plumb-line, covering agents.md — plumb-line, where things are written down, four things that are load-bearing, tracking (recursive-spine convention) and moments map.

slopstopper/plumb-line · 1,774 tokens

agentic-seo-skills AGENTS.md

Instructions for agencia-conversion/agentic-seo-skills, covering agentic seo — agent runtime contract, mission, six pillars, audience and delivery — web companion first.

agencia-conversion/agentic-seo-skills · 2,377 tokens

superpowers-graph CLAUDE.md

Claude Code instructions for RonMizrahi/superpowers-graph, covering superpowers-graph — repository guide, sources of truth, changing the graph — the eval ships with the change and rules that are not obvious from the code.

RonMizrahi/superpowers-graph · 738 tokens