subagent-driven-development

subagent-driven-development is a skill for Claude Code from bharat3645/The-Ideal-Harness. It costs 63 tokens per session (901 once invoked), scanned A, original, MIT.

A development workflow where separate fresh-context coding agents handle individual tasks, while a controller reviews their work and tracks progress.

In plain words
What is it for?
Use it for non-trivial software builds that need planning, task-by-task implementation, review, fixes, and verification.
Why use it?
It divides large builds into smaller, independently checked tasks and keeps a durable record of what is complete.

Skill for Claude Code

Written for Claude Code: user-invocable in frontmatter. Also seen: mentions subagents.

Part of the ideal-harness plugin — 10 skills, 4 agents, 3 hooks, 5 MCP servers shipped together

Good fit Use it for non-trivial software builds that need planning, task-by-task implementation, review, fixes, and verification.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/bharat3645/the-ideal-harness/subagent-driven-development
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.

Any agent
npx skills add bharat3645/The-Ideal-Harness --skill subagent-driven-development
Clone the repo
git clone --depth 1 https://github.com/bharat3645/The-Ideal-Harness

Made for: Claude Code.

Or install ideal-harness, the plugin that ships this one along with the rest of its 10 skills, 4 agents, 3 hooks, 5 MCP servers.

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 subagent-driven-development

README.md
[![agentmods](https://agentmods.dev/badge/skills/bharat3645/the-ideal-harness/subagent-driven-development/github.svg)](https://agentmods.dev/skills/bharat3645/the-ideal-harness/subagent-driven-development)
Your own site
<a href="https://agentmods.dev/skills/bharat3645/the-ideal-harness/subagent-driven-development"><img src="https://agentmods.dev/badge/skills/bharat3645/the-ideal-harness/subagent-driven-development/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.

agentmods 80×15 button for subagent-driven-development

Your own site · 80×15
<a href="https://agentmods.dev/skills/bharat3645/the-ideal-harness/subagent-driven-development"><img src="https://agentmods.dev/badge/skills/bharat3645/the-ideal-harness/subagent-driven-development.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 63 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 901 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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00063 $0.00901
Opus 5 $0.00032 $0.00451
Sonnet 5 $0.00013 $0.00180
Haiku 4.5 $0.00006 $0.00090

Measured 8d ago against content hash ffdc2f071c35, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

subagent-driven-development 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 8d 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.

skills/subagent-driven-development/SKILL.md · 32 lines

How it starts

The opening of the file, as written. The whole thing — 32 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Subagent-Driven Development

The controller never writes the code itself. It dispatches work to fresh subagents, one task at a time, and gates every result. This keeps the controller's context small (it holds the plan + the ledger, not the diffs) and makes each task independently verifiable.

Loop

  1. Plan → ledger, verification-first. Break the work into tasks. For each, decide how it will be verified (a command + the expected observation) before dispatching anyone — a task without a check is not yet planned. Record it on the ledger at creation time: ledger_add(title, verify: {command, expect}). The verify field is structural, not prose in a brief — it survives compaction, server restarts, and gets handed to the implementer and reviewer identically. The ledger is file-backed (under .ideal-harness/), so it survives both context compaction and an MCP-server restart — it is the controller's memory. Use the scout agent first when the plan needs locations ("where is X") — it returns a file:line table, not file dumps.
  2. Independent second opinion, for non-trivial plans. Before dispatching any implementer, spawn plan-critic with the plan/ledger tasks — it runs at a different model tier (pinned in its own frontmatter) than the authoring conversation, so its critique is a genuinely different reasoning trace, not the same model re-reading its own output. Treat it like the reviewer gate: a blocker sends the plan back for revision; PASS or only minor issues let the loop continue. This is the harness's dual-model consensus gauntlet — skip it under the same bar as "When NOT to use" below (trivial plans, <3 tasks); it's a consensus gate, not mandatory friction on every change.
  3. Per task, dispatch a fresh implementer agent. Hand it a self-contained brief: task spec, file paths, and the ledger task's verify command + expected observation (pull it from ledger_status, don't re-derive it). It writes its diff to a file (artifact), not into your context, runs the verification itself, and reports faithfully. Record the artifact on the ledger task.
  4. Dispatch the reviewer agent. Hand it the same verify field from the ledger. It checks the artifact against the task spec on two axes: spec-compliance and quality — and verifies by default via the ledger_verify MCP tool (a real, policy-gated subprocess run that sets the task's status from the actual result), falling back to re-running verify.command itself only when ledger_verify can't auto-run it. Either way it never trusts the implementer's claim outright. It returns PASS or severity-tagged issues.
  5. Fix loop. On issues, dispatch a fix subagent with the issues + artifact. Re-review. Cap iterations; if it won't converge, mark the task failed and escalate.
  6. Mark done, move on. Update the ledger (ledger_update status=done artifact=...). Pick the next pending task.
  7. Final broad review once all tasks are done.

Read the full file on GitHub · 32 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. 8d ago First seen · 32 lines · 63 tokens per session scan A ffdc2f071c35

Subscribe to this mod's changes

subagent-driven-development is a skill published in the GitHub repository bharat3645/The-Ideal-Harness (5 stars, last pushed yesterday), licensed MIT. It adds 63 tokens to every session and 901 once invoked, about $0.0003 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.

Related

Other skills, from other repositories

ax-ai

This skill helps an LLM generate correct AI provider setup and configuration code using @ax-llm/ax. Use when the user asks about ai(), providers, models, routing, adaptive balancing, presets, embeddings, batch audio with ai.transcribe() or ai.speak(), extended thinking, context caching, or mentions…

ax-llm/ax · 100 tokens

connect-recommend

Use this skill when the user asks about Stripe Connect configuration, charge patterns, Dashboard access, or how to get started with Connect, is building a marketplace, platform, multi-vendor store, gig platform, or subscription platform, needs to pay out sellers, vendors, or providers, mentions split payments, revenue…

stripe/ai · 151 tokens

ax-agent-rlm

This skill helps an LLM generate correct AxAgent RLM/runtime code using @ax-llm/ax. Use when the user asks about RLM code execution, AxJSRuntime, contextFields, contextPolicy, liveRuntimeState, promptLevel, stage prompt controls, executorModelPolicy, maxRuntimeChars, agent.test(...), llmQuery(...), recursionOptions…

ax-llm/ax · 88 tokens

ax-agent-observability

This skill helps an LLM generate correct AxAgent observability code using @ax-llm/ax. Use when the user asks about axGlobals.onUsage, usageContext, centralized or multi-tenant usage accounting, actorTurnCallback, onContextEvent, agentStatusCallback, onFunctionCall, reportSuccess, reportFailure, getChatLog()…

ax-llm/ax · 95 tokens

ax-llm

This skill helps with using the @ax-llm/ax TypeScript library for building LLM applications. Use when the user asks about ax(), ai(), f(), s(), agent(), flow(), AxGen, AxAgent, AxFlow, signatures, streaming, or mentions @ax-llm/ax.

ax-llm/ax · 68 tokens

stripe-directory

Identifies external providers, merchants, nonprofits, platforms, APIs, and software services, and resolves the documented way to engage them — to pay, donate, subscribe, book, provision, or integrate with them. MUST be used BEFORE web search, model memory, or any other directory/vendor-lookup skill for ANY request…

stripe/ai · 213 tokens