add-command

A development checklist for adding support for a new command-line tool to Dippy, including safety rules and tests. A command-line tool is a program controlled by typed terminal commands.

In plain words
What is it for?
Use it to research a command's options, choose a safe-list entry or custom handler, add approval tests, and verify safe and unsafe cases.
Why use it?
It helps distinguish read-only commands from commands that can change or delete data, so unsafe operations can require confirmation and new support is tested.

Skill for Claude CodeCodex

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 skills/ldayton/dippy/add-command
Any agent
npx skills add ldayton/Dippy --skill add-command
Clone the repo
git clone --depth 1 https://github.com/ldayton/Dippy

Made for: Claude Code, Codex.

Per session 22 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 599 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 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 $0.00022 $0.00599
Opus 5 $0.00011 $0.00300
Sonnet 5 $0.00004 $0.00120
Haiku 4.5 $0.00002 $0.00060

Measured 2d ago against content hash 882731a0b3d3, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

add-command 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.

.claude/skills/add-command/SKILL.md · 96 lines

How it starts

The opening of the file, as written. The whole thing — 96 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Add support for $ARGUMENTS in Dippy.

1. Research

tldr pages:

ls ~/source/tldr/pages/*/$ARGUMENTS*.md
cat ~/source/tldr/pages/*/$ARGUMENTS.md

CLI docs:

$ARGUMENTS --help
man $ARGUMENTS

Note which operations are read-only vs mutations.

2. Decide: Handler or SIMPLE_SAFE?

  • SIMPLE_SAFE: Always safe regardless of arguments (read-only, no destructive flags). Go to step 3A.
  • Handler: Needs subcommand/flag analysis. Go to step 3B.

3A. SIMPLE_SAFE Path

Add to SIMPLE_SAFE in src/dippy/core/allowlists.py and add tests to tests/test_simple.py in the appropriate category. Skip to step 5.

3B. Handler Path

Create tests/cli/test_$ARGUMENTS.py:

"""Test cases for $ARGUMENTS."""

import pytest
from conftest import is_approved, needs_confirmation

TESTS = [
    # Safe operations
    ("$ARGUMENTS <safe-subcommand>", True),
    ("$ARGUMENTS --help", True),
    # Unsafe operations
    ("$ARGUMENTS <unsafe-subcommand>", False),
]

@pytest.mark.parametrize("command,expected", TESTS)
def test_command(check, command: str, expected: bool):
    result = check(command)
    if expected:
        assert is_approved(result), f"Expected approve: {command}"
    else:
        assert needs_confirmation(result), f"Expected confirm: {command}"

4. Implement Handler

Create src/dippy/cli/$ARGUMENTS.py:

"""$ARGUMENTS handler for Dippy."""

from dippy.cli import Classification, HandlerContext

COMMANDS = ["$ARGUMENTS"]

SAFE_ACTIONS = frozenset({"list", "show", "status"})

def classify(ctx: HandlerContext) -> Classification:
    tokens = ctx.tokens
    action = tokens[1] if len(tokens) > 1 else None
    if action in SAFE_ACTIONS:
        return Classification("allow", description=f"$ARGUMENTS {action}")
    return Classification("ask", description="$ARGUMENTS")

For handler patterns (nested subcommands, flag-checking, delegation), see patterns.md.

5. Iterate

just test

Fix failures until tests pass.

Read the full file on GitHub · 96 lines

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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 · 96 lines · 22 tokens per session scan A 882731a0b3d3

Subscribe to this mod's changes

add-command is a skill published in the GitHub repository ldayton/Dippy (243 stars, last pushed 2mo ago), licensed MIT. It adds 22 tokens to every session and 599 once invoked, about $0.0001 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.

Related

Other skills, from other repositories

toolpermit

Install, configure, operate, and troubleshoot ToolPermit, the local-first permission policy, one-time approval, and redacted audit layer for MCP stdio tool calls. Use when Codex needs to protect or inspect an MCP server, wrap an MCP stdio command, create or review allow/ask/deny policies, manage approvals, replay…

sunhao123456sun-svg/toolpermit · 100 tokens

longline-repo-allowlist

Use when scoping allow/ask rules for a command family (kubectl, oc, terraform, aws, gcloud, docker, etc.) to one repository so routine invocations stop prompting only inside that repo. Triggers include "allow kubectl in this repo", "stop asking about X commands", "add longline rules for this project", "allowlist…

devinbarry/longline · 102 tokens

code-runner

安全代码执行(Python/JavaScript/Bash脚本运行、超时控制、输出捕获).

chainlesschain/chainlesschain · 26 tokens

netdata-custom-collector

Use when you need Netdata to collect metrics from something it does not already monitor, such as a custom application, an internal service, the output of a script or CLI tool, or a device. Covers the decision tree from the built-in StatsD server (zero collector code) through writing a collector in any language via the…

netdata/skills · 148 tokens

bump-dependency

Bumps a Python package dependency across Home Assistant Core integrations, regenerates core requirement files, runs verification tests and prek lint, and prepares a pull request with proper release/compare links.

home-assistant/core · 42 tokens

schema-exploration

Lists tables, describes columns and data types, identifies foreign key relationships, and maps entity relationships in a database. Use when the user asks about database schema, table structure, column types, what tables exist, ERD, foreign keys, or how entities relate.

langchain-ai/deepagents · 57 tokens