task-new

task-new is a skill for Claude Code, Codex from loncadev/baron. It costs 82 tokens per session (888 once invoked), scanned A, original, Apache-2.0.

Project instructions for developing Baron, a TypeScript work-orchestration layer, including its architecture, tools, coding conventions, and test commands. A monorepo is a repository containing multiple related packages managed together.

In plain words
What is it for?
Guide Baron development with its TypeScript, Node.js, pnpm, Vitest, Biome, and build conventions, including how to test, type-check, lint, and package the code.
Why use it?
They give coding agents the project's technical rules and the commands needed to check changes, reducing inconsistent implementations and missed verification.

Skill for Claude CodeCodex

Part of the baron plugin — 10 skills shipped together

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/loncadev/baron/task-new
Any agent
npx skills add loncadev/baron --skill task-new
Clone the repo
git clone --depth 1 https://github.com/loncadev/baron

Made for: Claude Code, Codex.

Or install baron, the plugin that ships this one along with the rest of its 10 skills.

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 task-new

README.md
[![agentmods](https://agentmods.dev/badge/skills/loncadev/baron/task-new.svg)](https://agentmods.dev/skills/loncadev/baron/task-new)
Your own site
<a href="https://agentmods.dev/skills/loncadev/baron/task-new"><img src="https://agentmods.dev/badge/skills/loncadev/baron/task-new.svg" alt="Measured on agentmods" height="20"></a>
Per session 82 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 888 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 $0.00082 $0.00888
Opus 5 $0.00041 $0.00444
Sonnet 5 $0.00016 $0.00178
Haiku 4.5 $0.00008 $0.00089

Measured 3d ago against content hash 611a5f279df9, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

task-new 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 3d 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/claude-code/skills/task-new/SKILL.md · 63 lines

How it starts

The opening of the file, as written. The whole thing — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Create a work item

Turn the user's request into a work item that is actually actionable by whoever picks it up — not a bare title — then run the task-new recipe as ONE deterministic call. Creation only; it does not branch or change your checkout (that's /task-start).

Steps

  1. Infer the type role from the wording: defect/broken/error → bug; a user-facing capability → story; a concrete engineering step → task. If ambiguous, ask via AskUserQuestion (task / bug / story).

  2. Reach a well-specified item BEFORE creating — don't open a hollow title. A title-only issue is guesswork for the next person. Decide how much you already have:

    • The context already settles it — the conversation made the decisions and now says "make a task for it", or the user's message already carries the what + why (+ repro for a bug). Then synthesize the body from that context; don't re-ask what you already know.
    • The input is thin — just a short title, no substance. Then grill to decision-saturation with AskUserQuestion before creating: ask only the few questions that actually change the body — typically what problem / goal, why now / value, what "done" means (acceptance), and for a bug the repro → expected → observed. Stop as soon as you could write a body someone else could act on; don't interrogate past that.
    • Respect a deliberate quick capture. If the user says "just create it" / "only the title for now", create with what you have — a note in the body that detail is TBD, not a wall of questions.

    Compose the gathered detail into a clear body (markdown). This is what makes the item useful; the title alone never is.

  3. Resolve the parent (a dangling item gets lost):

    • bug/task belong under a story; a story belongs under an epic.
    • Find candidates with baron_issue_read op=query (typeRole: "story" or "epic") and offer the best 2-3 via AskUserQuestion, plus "no parent". Never invent a parent id.
    • On providers without native hierarchy the parent is emulated/degraded per policy — Baron's job.

Read the full file on GitHub · 63 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. 3d ago First seen · 63 lines · 82 tokens per session scan A 611a5f279df9

Subscribe to this mod's changes

task-new is a skill published in the GitHub repository loncadev/baron (1 stars, last pushed 13d ago), licensed Apache-2.0. It adds 82 tokens to every session and 888 once invoked, about $0.0004 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 2026-08-31.