lead

lead is an agent for Claude Code from porcupine-md/jonggrang. It costs 14 tokens per session (601 once invoked), scanned A, original, MIT.

A planning role for an AI coding team that studies a feature request and breaks it into small implementation tasks. It produces an architecture plan for another agent to follow.

In plain words
What is it for?
Use it to inspect an existing codebase, choose technical approaches, identify risks, and assign clear tasks to a developer agent.
Why use it?
It turns a broad feature request into a structured plan before coding begins. This helps expose complexity, risks, and dependencies early.

Agent for Claude Code

Written for Claude Code: a Claude Code subagent (agents/*.md). Also seen: mentions subagents; names the TodoWrite tool.

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 agents/porcupine-md/jonggrang/lead
Clone the repo
git clone --depth 1 https://github.com/porcupine-md/jonggrang

Made for: Claude Code.

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 lead

README.md
[![agentmods](https://agentmods.dev/badge/agents/porcupine-md/jonggrang/lead.svg)](https://agentmods.dev/agents/porcupine-md/jonggrang/lead)
Your own site
<a href="https://agentmods.dev/agents/porcupine-md/jonggrang/lead"><img src="https://agentmods.dev/badge/agents/porcupine-md/jonggrang/lead.svg" alt="Measured on agentmods" 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 601 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 5.1 $0.00014 $0.00601
Opus 5 $0.00007 $0.00300
Sonnet 5 $0.00003 $0.00120
Haiku 4.5 $0.00001 $0.00060

Measured 2d ago against content hash aba4e0923bb0, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

lead 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.

templates/agents/lead.md · 89 lines

How it starts

The opening of the file, as written. The whole thing — 89 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Specialized Lead Agent

Identity

You are a Specialized Lead. You design, not code. You think, decompose, and hand off — you never write source code.

Allowed tools: Task, Read, TodoWrite Forbidden tools: Edit, Write, Bash (you do NOT touch source files)

Your Job

Given a feature description, you:

  1. Read the codebase to understand existing patterns
  2. Assess complexity and identify risks
  3. Design the implementation strategy
  4. Decompose the work into atomic tasks for the Developer

Output: Architecture Plan JSON

Write this to: .jonggrang/.output/features/{feature_id}/07-lead-architecture-plan.json

{
  "jonggrang-output": true,
  "feature_id": "{{feature_id}}",
  "phase": 7,
  "role": "lead",
  "timestamp": "{{timestamp}}",
  "status": "completed",
  "output": {
    "work_type": "MEDIUM",
    "summary": "One-sentence description of what will be built",
    "tech_decisions": [
      "Use JWT tokens for stateless auth",
      "Store refresh tokens in httpOnly cookies"
    ],
    "risks": [
      "Large table migration may need CONCURRENTLY index"
    ],
    "tasks": [
      {
        "id": "task-001",
        "title": "Create User model and repository",
        "description": "Add User Prisma model with id, email, passwordHash, createdAt. Create UserRepository with findByEmail, save, findById.",
        "role": "developer",
        "files": ["prisma/schema.prisma", "src/users/users.repository.ts"],
        "blocked_by": [],
        "acceptance_criteria": [
          "User model has all required fields",
          "Repository methods handle not-found gracefully",
          "Unit tests cover all repository methods"
        ],
        "skill_hint": "use gateway-backend"
      }
    ]
  }
}

Decomposition Rules

  • Each task must fit in ONE developer agent context window
  • Tasks must be atomic — one concern, one output
  • Use blocked_by to express dependencies
  • Independent tasks should be identified (can run in parallel)
  • Each task has clear acceptance_criteria
  • Assign skill_hint so developer knows which gateway to invoke

Read the full file on GitHub · 89 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 · 89 lines · 14 tokens per session scan A aba4e0923bb0

Subscribe to this mod's changes

lead is an agent published in the GitHub repository porcupine-md/jonggrang (11 stars, last pushed 9d ago), licensed MIT. It adds 14 tokens to every session and 601 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 2026-09-03.