init-helpers

A project setup command that detects the codebase's programming stack and drafts a .claude/helpers.json configuration file. The file tells related development commands which tests, tools, and project conventions to use.

In plain words
What is it for?
Initialising development-helper settings in a repository or refreshing them after the project's stack changes.
Why use it?
It removes the need to configure those project-specific commands by hand at the start of work on a repository.

Command

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 commands/refo/claude-dev-helpers/init-helpers
Clone the repo
git clone --depth 1 https://github.com/refo/claude-dev-helpers
Per session 41 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,971 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 2 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 $0.00041 $0.01971
Opus 5 $0.00020 $0.00986
Sonnet 5 $0.00008 $0.00394
Haiku 4.5 $0.00004 $0.00197

Measured 2d ago against content hash b755dedc4f9b, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

init-helpers scanned grade C with 2 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.

Reads agent configuration directoriesmediumAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

- Existing helpers config: !`cat .claude/helpers.json 2>/dev/null || echo "(none — this is a fresh project)"`

Unrestricted tool accessmediumExcessive agency

A wildcard tool grant or "run any command" leaves no least-privilege boundary at all.

You don't need to run any commands beyond reading files. Use `Read`, `Glob`, and `Grep`. Reach for `WebSearch` only if you encounter an exotic stack you genuinely don't recognize from training.
plugins/dev-helpers/commands/init-helpers.md · 133 lines

How it starts

The opening of the file, as written. The whole thing — 133 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Init Helpers

Context

  • Existing helpers config: !cat .claude/helpers.json 2>/dev/null || echo "(none — this is a fresh project)"
  • Working directory: !pwd
  • Top-level files: !ls -1
  • Git remote (if any): !git remote get-url origin 2>/dev/null || echo "(no git remote)"
  • Top of CLAUDE.md (if any): !head -40 CLAUDE.md 2>/dev/null || echo "(no CLAUDE.md)"

Purpose

This command bootstraps <repo>/.claude/helpers.json — the per-project config file that the rest of the dev-helpers plugin commands (commit, next-issue, tdd, setup-pre-commit, etc.) read to tailor their behavior to this specific project.

You only need to run it once per repo, after installing the dev-helpers plugin. Re-run it if the project's stack changes (e.g. switched package manager, added a new linter, moved issue trackers).

What you (the assistant) must do

1. Bail out gracefully if helpers.json already exists

If .claude/helpers.json is already present in the project (see Context above):

  • Show the user the current contents.
  • Ask: "A helpers.json already exists. Do you want to (a) leave it alone, (b) review and update fields, or (c) regenerate from scratch?"
  • Wait for their answer. Do NOT overwrite without explicit confirmation.

2. Detect the stack from project files

You have full access to the working directory. Inspect it thoroughly:

  • Manifests and lockfiles to identify language + package manager: package.json, bun.lock, bun.lockb, pnpm-lock.yaml, yarn.lock, package-lock.json, pyproject.toml, uv.lock, poetry.lock, requirements.txt, Pipfile, go.mod, Cargo.toml, Gemfile, mix.exs, composer.json, pubspec.yaml, build.gradle*, pom.xml, *.csproj, *.fsproj, *.sln, flake.nix, Justfile, Makefile.
  • Linter / formatter configs: biome.json, .eslintrc*, .prettierrc*, ruff.toml, .rubocop.yml, rustfmt.toml, clippy.toml, .editorconfig.
  • Test configs: vitest.config.*, jest.config.*, playwright.config.*, pytest.ini, tox.ini, presence of *.test.ts, test_*.py, *_test.go, spec/ directory, etc.
  • Pre-commit / git-hook configs: .husky/, .pre-commit-config.yaml, lefthook.yml/.yaml, .git/hooks/pre-commit.
  • Issue tracker hints: git remote host (GitHub / GitLab / Bitbucket), .github/ directory, references in CLAUDE.md.
  • CLAUDE.md for any project-specific rules already documented (commit scope conventions, "don't run X" rules, naming conventions).
  • package.json scripts section (or equivalent) for test, typecheck, lint, format, dev, build script names.

Read the full file on GitHub · 133 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. 2d ago First seen · 133 lines · 41 tokens per session scan C b755dedc4f9b

Subscribe to this mod's changes

init-helpers is a command published in the GitHub repository refo/claude-dev-helpers (2 stars, last pushed 2mo ago), licensed MIT. It adds 41 tokens to every session and 1,971 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it C with 2 findings (reads agent configuration directories, unrestricted tool access). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.