New: 98,000+ skills indexed from GitHub repos with 100+ stars · every skill safety-checkedBrowse top skillsTop skills
Esc to closeBrowse all skills
Submit a skillBrowse skills
Submit a skillBrowse skills

Agent skill

Subagent Driven Development

by HezaoHezao in HezaoHezao/poirot

Execute plans task-by-task with two-stage review. An agent skill from HezaoHezao/poirot.

MITAuto-check passedAgent Workflows

Install Subagent Driven Development

skills CLI
$ npx skills add HezaoHezao/poirot --skill subagent-driven-development -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install HezaoHezao/poirot subagent-driven-development --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/HezaoHezao/poirot.git skills-src && mkdir -p .claude/skills && cp -r skills-src/poirot/backend/agents/skill/builtin_skills/software-development/subagent-driven-development .claude/skills/subagent-driven-development && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
subagent-driven-development
GitHub stars
250
Token cost
~1.3k tokens
SKILL.md length
620 words
Files
1
Skills in repo
23
Repo updated
First seen
Licence
MIT

At a glance

Execute plans task-by-task with two-stage review. An agent skill from HezaoHezao/poirot.

  • Works in 5 steps: Read and Parse Plan → Per-Task Workflow → Handling Failures → …
  • Tasks that involve Subagents
  • SKILL.md covers Overview, When to Use, The Process and Two-Stage Review Detail, plus 1 more section
  • Calls git

What it does

Subagent Driven Development is an agent skill from HezaoHezao/poirot. Execute plans task-by-task with two-stage review.

Its SKILL.md is about 1.3k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Agent Workflows, covering Subagents and Planning. The repository describes itself as: Poirot is a deep research agent kernel built for those who care about how agents are architected. The licence is MIT.

When your agent uses it

  • Tasks that involve Subagents
  • Tasks that involve Planning

Example prompts

  • “/subagent-driven-development”

Requirements

  • Pre-approved tools (allowed-tools): bash, read_file, write_file, str_replace, list_dir

Workflow steps

5 steps, taken from the step headings in SKILL.md.

  1. Read and Parse Plan
  2. Per-Task Workflow
  3. Handling Failures
  4. Spec Compliance
  5. Code Quality

What it can do on your machine

Read from SKILL.md and the folder at commit 86bf279. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves these tools, so the agent can use them without asking each time:

    • bash
    • read_file
    • write_file
    • str_replace
    • list_dir

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • git

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use git, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Subagent Driven Development loads about 1.3k tokens when it runs. Until then it costs about 19 tokens; SKILL.md has 620 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~19
When it runs · the whole SKILL.md, loaded when a task matches
~1.3k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from HezaoHezao/poirot at commit 86bf279, republished under its MIT licence (© HezaoHezao). 620 words, ~1,285 tokens.

Download SKILL.mdSave it as .claude/skills/subagent-driven-development/SKILL.md (or your agent's skills folder).
name
subagent-driven-development
description
Execute plans task-by-task with two-stage review.
allowed-tools
bash, read_file, write_file, str_replace, list_dir
enabled
true
related-skills
plan, requesting-code-review, test-driven-development
license
MIT
author
Adapted from hermes-agent (Nous Research, MIT); obra/superpowers

Subagent-Driven Development

Overview

Execute implementation plans task-by-task with systematic two-stage review between tasks.

Core principle: Fresh focus per task + two-stage review (spec then quality) = high quality, fast iteration.

Poirot note: The original skill dispatches a fresh subagent per task via delegate_task. Poirot has no subagents, so this version executes tasks sequentially in the same context with a deliberate context-reset between tasks. The two-stage review methodology is preserved.

When to Use

  • You have an implementation plan (from the plan skill or user requirements)
  • Tasks are mostly independent
  • Quality and spec compliance are important
  • You want automated review between tasks

The Process

1. Read and Parse Plan

Read the plan file. Extract ALL tasks with full text and context upfront:

read_file(".poirot/plans/feature-plan.md")

Create a todo list with all tasks. Read the plan ONCE — don't re-read per task.

2. Per-Task Workflow

For EACH task in the plan:

Step 1: Context Reset

Before starting each task, deliberately reset your focus:

  • Re-read only the current task's description
  • Forget previous tasks' implementation details (they're committed)
  • Treat each task as if a fresh agent is picking it up
Step 2: Implement

Follow the task's steps exactly:

  1. Write the failing test (TDD — see test-driven-development skill)
  2. Run test to verify failure
  3. Write minimal implementation
  4. Run test to verify pass
  5. Run full test suite to check for regressions
  6. Commit
bash("pytest tests/test_feature.py::test_name -v")
# ... implement ...
bash("pytest tests/ -q")
bash("git add -A && git commit -m 'feat: task N description'")
Step 3: Stage 1 Review — Spec Compliance

After implementation, review against the plan:

  • Does the implementation match the task's stated objective?
  • Are all files listed in the task created/modified?
  • Are all tests listed in the task written?
  • Does the implementation do what the task says, not more, not less?

If spec compliance fails: Fix before proceeding to Stage 2.

Step 4: Stage 2 Review — Code Quality

Review the committed diff for quality:

  • No hardcoded secrets, credentials
  • Input validation on user-provided data
  • Error handling for I/O/network/DB
  • No debug print/console.log left behind
  • No commented-out code
  • Tests cover happy path + edge cases
  • Code follows project conventions

Use the requesting-code-review skill's checklist for thoroughness.

If quality review fails: Fix, re-commit, re-review.

Step 5: Proceed to Next Task

Only after both reviews pass:

  1. Mark current task complete in the todo list
  2. Move to next task
  3. Repeat from Step 1 (Context Reset)
Show full SKILL.md (248 more words)Show less
3. Handling Failures

Implementation fails (tests don't pass):

  1. Use systematic-debugging skill to find root cause
  2. Fix the root cause
  3. Re-run tests
  4. If 3+ fixes fail → question the approach, discuss with user

Review fails (spec or quality):

  1. Fix the specific issues found
  2. Re-commit
  3. Re-review only the fixed parts

Task is blocked (depends on unfinished work):

  1. Note the blocker
  2. Skip to next independent task
  3. Return to blocked task when dependency is resolved

Two-Stage Review Detail

Stage 1: Spec Compliance

Re-read the task from the plan. Compare against what you actually built:

Task says: "Create User model with email and password_hash fields"
Check:
- [ ] User model exists
- [ ] Has email field
- [ ] Has password_hash field
- [ ] No extra fields not in the task
- [ ] Test exists for the model

Common spec violations:

  • Implementing more than the task asks ("while I'm here" scope creep)
  • Missing a file the task lists
  • Different API shape than the task specifies
  • No test for the feature
Stage 2: Code Quality

Review the git diff for the current task:

bash("git diff HEAD~1 HEAD")

Check for security, correctness, and conventions. Use the requesting-code-review skill's security scan + self-review checklist.

Pitfalls

  • Context accumulation: without subagents, context grows across tasks. If you feel confused about which task you're on, re-read the plan + current task.
  • Scope creep: "while I'm here" edits to previous tasks' code. Don't. Each task is atomic.
  • Skipping review: the two-stage review IS the value of this skill. Without it, you're just implementing sequentially.
  • Re-reviewing everything: after a fix, only re-review the fixed parts, not the entire task.
  • Forgetting to commit: commit after each task. Don't accumulate uncommitted changes across tasks — if a later task breaks, you can't cleanly revert.

© HezaoHezao, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in poirot/backend/agents/skill/builtin_skills/software-development/subagent-driven-development of HezaoHezao/poirot.

Open the folder on GitHubat commit 86bf279

Compare with similar skills

Subagent Driven Development next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Subagent Driven Development compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Subagent Driven Development this skillHezaoHezao/poirot250—~1.3kAutomated safety check: PassMIT
Subagent Driven DevelopmentAsvarox/allkaraoke26138 repos~1.2kAutomated safety check: PassNone
Execumputun/cc-thingz484—~8kAutomated safety check: PassMIT
Executing PlansGanyuanRan/Aegis1.3k1 repos~2.3kAutomated safety check: PassMIT
Autopilot End-to-End Buildernick-vels/skills407—~2.5kAutomated safety check: NotesMIT
Workflow Orchestrationvxcozy/workflow-orchestration116—~1kAutomated safety check: PassMIT

Similar skills

  • Subagent Driven Development

    Asvarox/allkaraoke

    A skill your agent uses when executing implementation plans with independent tasks in the current session

    261 GitHub starsUsed in 38 repos~1.2k tokens
    Agent WorkflowsAuto-check passed

  • Exec

    umputun/cc-thingz

    Execute plan tasks sequentially using subagents. An agent skill from umputun/cc-thingz.

    484 GitHub stars~8k tokensUpdated 2 days ago
    Agent WorkflowsAuto-check passed

  • Executing Plans

    GanyuanRan/Aegis

    A skill your agent uses when executing a written implementation plan across sessions or with review checkpoints.

    1.3k GitHub starsUsed in 1 repo~2.3k tokens
    Agent WorkflowsAuto-check passed

  • Autopilot End-to-End Builder

    nick-vels/skills

    Builds an app, site, bot or feature from a spoken idea through requirements, spec, tickets, subagent work, review and acceptance, with a live progress dashboard.

    407 GitHub stars~2.5k tokensUpdated 5 days ago
    Agent WorkflowsAuto-check: notes

  • Workflow Orchestration

    vxcozy/workflow-orchestration

    Disciplined task execution with planning, verification, and self-improvement loops.

    116 GitHub stars~1k tokensUpdated 5 mo ago
    Agent WorkflowsAuto-check passed

  • Blueprint

    byungjunjang/jangpm-meta-skills

    Agentic system design blueprint generator. An agent skill from byungjunjang/jangpm-meta-skills.

    121 GitHub stars~2k tokensUpdated 2 mo ago
    Agent WorkflowsAuto-check passed

More from HezaoHezao/poirot

All 23 skills in this repo

  • Requesting Code Review

    HezaoHezao/poirot

    Pre-commit review: security scan, quality gates, auto-fix. An agent skill from HezaoHezao/poirot.

    250 GitHub starsUsed in 5 repos~1.6k tokens
    Auto-check passed

  • Research Paper Writing

    HezaoHezao/poirot

    ML paper pipeline: experiment design to submission. An agent skill from HezaoHezao/poirot.

    250 GitHub starsUsed in 1 repo~1.4k tokens
    Auto-check passed

  • Skill Creator

    HezaoHezao/poirot

    Create, edit, and evaluate agent skills iteratively. An agent skill from HezaoHezao/poirot.

    250 GitHub starsUsed in 1 repo~838 tokens
    Auto-check passed

  • Academic Paper Review

    HezaoHezao/poirot

    Structured peer-review of academic papers. An agent skill from HezaoHezao/poirot.

    250 GitHub stars~1.8k tokensUpdated 2 mo ago
    Auto-check passed

  • Blogwatcher

    HezaoHezao/poirot

    Monitor blogs and RSS/Atom feeds via blogwatcher-cli. An agent skill from HezaoHezao/poirot.

    250 GitHub stars~854 tokensUpdated 2 mo ago
    Auto-check passed

  • Bootstrap

    HezaoHezao/poirot

    Onboarding conversation to generate a user profile. An agent skill from HezaoHezao/poirot.

    250 GitHub stars~1.2k tokensUpdated 2 mo ago
    Auto-check passed

Categories

Questions about Subagent Driven Development

What does Subagent Driven Development do?

Execute plans task-by-task with two-stage review. An agent skill from HezaoHezao/poirot. Subagent Driven Development is an agent skill from HezaoHezao/poirot. Execute plans task-by-task with two-stage review.

When should I use Subagent Driven Development?

Subagent Driven Development fits situations like: tasks that involve Subagents; tasks that involve Planning.

How do I install Subagent Driven Development in Claude Code?

Run `npx skills add HezaoHezao/poirot --skill subagent-driven-development -a claude-code`. Or copy the skill folder (poirot/backend/agents/skill/builtin_skills/software-development/subagent-driven-development in HezaoHezao/poirot) into .claude/skills/subagent-driven-development in your project. Claude Code loads it when a task matches its description.

How do I install Subagent Driven Development in Codex?

Run `npx skills add HezaoHezao/poirot --skill subagent-driven-development -a codex`. Or copy the skill folder (poirot/backend/agents/skill/builtin_skills/software-development/subagent-driven-development in HezaoHezao/poirot) into .agents/skills/subagent-driven-development in your project. Codex loads it when a task matches its description.

Can I use Subagent Driven Development in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add HezaoHezao/poirot --skill subagent-driven-development -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/subagent-driven-development, .gemini/skills/subagent-driven-development, .github/skills/subagent-driven-development and .opencode/skills/subagent-driven-development in your project.

What does Subagent Driven Development need to run?

Going by SKILL.md and its folder, Subagent Driven Development needs the command-line tools its instructions call (git). Its frontmatter pre-approves these tools: bash, read_file, write_file, str_replace, list_dir.

Does Subagent Driven Development access the network?

SKILL.md contains no URLs. Its commands use git, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Subagent Driven Development safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Subagent Driven Development use?

Subagent Driven Development is published under the MIT licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Subagent Driven Development use?

About 1.3k tokens (SKILL.md is roughly 5.1k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Subagent Driven Development?

Skills that share tags, products or a category with Subagent Driven Development: Subagent Driven Development (Asvarox/allkaraoke, 261 stars), Exec (umputun/cc-thingz, 484 stars), Executing Plans (GanyuanRan/Aegis, 1.3k stars) and Autopilot End-to-End Builder (nick-vels/skills, 407 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Subagent Driven Development?

HezaoHezao (a GitHub user) maintains it in HezaoHezao/poirot, which has 250 GitHub stars. The repository holds 23 skills in this directory. The repository was last updated on July 28, 2026.

Source: HezaoHezao/poirot on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.