A command that starts a supervised work loop from a backlog, which is a list of tasks to complete. It validates the list, checks the Git working tree, switches to a separate branch, and begins with the first task.

Use it to run a planned batch of development tasks in a Git repository while the session supervises progress.

It removes the need to manually feed tasks to an agent one by one. The clean-tree and branch checks also reduce the chance of mixing this work with unrelated changes.

start — Command for Claude Code by krzysztofsurdy/code-virtuoso

Install

Getting it into your agent

This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.

Claude Code

/plugin marketplace add krzysztofsurdy/code-virtuoso

Claude Code

/plugin install loop-virtuoso

Made for: Claude Code.

Or install loop-virtuoso, the plugin that ships this one along with the rest of its 2 skills, 4 commands, 2 hooks.

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.

README.md

[![agentmods](https://agentmods.dev/badge/commands/krzysztofsurdy/code-virtuoso/start/github.svg)](https://agentmods.dev/commands/krzysztofsurdy/code-virtuoso/start)

Your own site

<a href="https://agentmods.dev/commands/krzysztofsurdy/code-virtuoso/start"><img src="https://agentmods.dev/badge/commands/krzysztofsurdy/code-virtuoso/start/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/commands/krzysztofsurdy/code-virtuoso/start"><img src="https://agentmods.dev/badge/commands/krzysztofsurdy/code-virtuoso/start.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 14 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 560 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

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.

Model	Per session	Once invoked
Fable 5.1	$0.00014	$0.00560
Opus 5.5	$0.00006	$0.00224
Sonnet 5.5	$0.00003	$0.00112
Haiku 4.5	$0.00001	$0.00056

Measured 2d ago against content hash 18307df05fa8, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

start 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.

plugins/loop-virtuoso/commands/start.md · 26 lines

What it actually says

Start the delivery loop (Mode A)

Run the setup script, passing along the backlog path argument if the user gave one:

${CLAUDE_PLUGIN_ROOT}/scripts/setup-start.sh $ARGUMENTS

No argument defaults to .code-virtuoso/loop/backlog.json at the git root, so it works from any subdirectory. If the project still has the old .delivery-loop/ directory, the script first moves it to .code-virtuoso/loop/ and copies an old global teams file to the user pipelines path; relay what it prints about that, including any advice to remove a .delivery-loop line from .git/info/exclude. The script validates the backlog, then refuses to start unless the working tree is clean (outside .code-virtuoso/) and backlog.branch is not the default branch. It switches to backlog.branch (creating it if needed), writes <git root>/.code-virtuoso/loop/state/session.local.json (the file that arms the hooks), and prints the first item's prompt. It does not archive anything: the backlog-compiler skill archives a finished or abandoned run before writing a new backlog. When the run ends with complete, suggest the finishing-branch skill.

If the script refuses, relay its message to the user: they need to commit or stash their changes, or pick a feature branch, before trying again.

After it runs:

Work the prompt the script printed - implement that one item and make its verifyCommand pass. Do not edit test files, the verify command, or anything under .code-virtuoso/; changes to those are rejected.
The Stop hook now intercepts each turn-end. It runs the verification gate on the item you just worked, records the result, and either feeds back the next item's prompt or lets the session end at a documented stop reason (complete, blocked_out, max_iterations, stalled, violation_limit, corrupted).
Do not decide an item or the backlog is done - only the gate flips an item to verified. Keep working each prompt the hook hands you until it releases the turn.

To stop early: /loop-virtuoso:cancel. To check progress: /loop-virtuoso:status.

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.

2d ago First seen · 26 lines · 14 tokens per session scan A 18307df05fa8

Subscribe to this mod's changes

start is a command published in the GitHub repository krzysztofsurdy/code-virtuoso (36 stars, last pushed yesterday), licensed MIT. It adds 14 tokens to every session and 560 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-10-22.

Claude Code Git and workflow Project management agent-skills ai-agents mcp plugins skills

start

Getting it into your agent

Wrote this? Show the measurements

What it costs to keep this loaded

Grade A, and why

Nothing flagged

What it actually says

Start the delivery loop (Mode A)

What this file has done since we first saw it

Other commands, from other repositories

flow-start

project

generate-changelog

update-deps

flow-release

amend