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/coderabbitai/git-worktree-runner/shgit clone --depth 1 https://github.com/coderabbitai/git-worktree-runnerWhat 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 | $0.00450 | $0.00450 |
| Opus 5 | $0.00225 | $0.00225 |
| Sonnet 5 | $0.00090 | $0.00090 |
| Haiku 4.5 | $0.00045 | $0.00045 |
Grade A, and why
git-worktree-runner sh.instructions.md scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- Keep dependencies minimal: only `git`, `sed`, `awk`, `find`, `grep` (avoid jq/curl unless justified). What it actually says
Shell Instructions
Architecture
Key Pattern: Everything is sourced at startup (set -e enabled). Functions call each other directly. No subshells except for hooks and AI tools.
Code Conventions
Shell Script Style
- Bash 3.2+ compatible (macOS default), but 4.0+ features allowed where appropriate
- Always quote variables:
"$var"not$var - Function-scoped vars: Use
local var="value" - Error handling: Check return codes; functions return 1 on failure
- Naming:
snake_casefor functions/vars,UPPER_CASEfor constants
Strict Mode & Safety
- Global
set -einbin/gtr: guard non-critical commands with|| true. - Prefer
[ ]over[[ ]]for POSIX portability (use[[only when needed). - Always quote glob inputs; disable unintended globbing (
set -ftemporarily if required).
Portability
- Target Bash 3.2+: avoid associative arrays; use simple string/loop constructs.
- Avoid
readarrayand process substitution unsupported in older Bash.
Debugging
- Quick trace:
bash -x ./bin/gtr <cmd>. - Inline: wrap suspicious block with
set -x/set +x. - Function presence:
declare -f create_worktreeordeclare -f resolve_target. - Variable inspection:
echo "DEBUG var=$var" >&2(stderr keeps stdout clean for command substitution).
External Commands
- Keep dependencies minimal: only
git,sed,awk,find,grep(avoid jq/curl unless justified). - Check availability before use if adding new tools.
Quoting & Paths
- Use
"${var}"; for loop over lines:while IFS= read -r line; do ... doneto preserve spaces. - Sanitize branch names via
sanitize_branch_name(do NOT duplicate logic elsewhere).
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.
- yesterday First seen · 48 lines · 450 tokens per session scan A c8f54135f6a0
git-worktree-runner sh.instructions.md is an instructions file published in the GitHub repository coderabbitai/git-worktree-runner (1,754 stars, last pushed 4d ago), licensed Apache-2.0. It adds 450 tokens to every session, about $0.0022 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
mulmoterminal CLAUDE.md
Instructions for receptron/mulmoterminal, covering claude.md — mulmoterminal, stack & package manager, run after changes, import a component at module scope, never inside a test and no emojis.
operator-oss CLAUDE.md
Instructions for iishyfishyy/operator-oss, covering claude.md, commands, architecture, the turn lifecycle (core flow) and key modules (by responsibility).
vibe-tree CLAUDE.md
Instructions for sahithvibudhi/vibe-tree, covering claude instructions, github actions failed build logs, electron app startup issues, launch with project and pull requests.
CommandMate CLAUDE.md
Instructions for Kewton/CommandMate, covering claude.md, ⚠️ 重要: モジュール詳細を claude.md に書かないこと, プロジェクト概要, 基本情報 and 技術スタック.
vscode-ext-tmux-worktree AGENTS.md
Instructions for kargnas/vscode-ext-tmux-worktree, covering ai agent guidelines, instruction, tested vs code, 1. codebase understanding and project structure.
CommandMate AGENTS.md
Instructions for Kewton/CommandMate, covering agents.md, scope, project context, working principles and coding rules.