Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsBlogPro
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges
  • Chrome Extension
  • Skill Manager

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Feature Development

ASecurity

DEFAULT WORKFLOW for ANY code change. Orchestrates the full rigorous development lifecycle: brainstorming → think-twice → debate → progressive-harsh-review → plan-and-execute → progressive-harsh-review → commit. This fires AUTOMATICALLY for code changes unless the user explicitly opts out. Sequences existing skills so no phase is skipped.

8 stars
0 votes
0 copies
0 views
Added 10/6/2026
ai-agentsgobashdebuggingcode-reviewgitdocumentation

Works with

mcp

Security Analysis

A100/100

Scanned 10/6/2026

$npx -y skills add bordenet/superpowers-plus --skill feature-development --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Feature Development?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Feature Development
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/bordenet-feature-development/badge)](https://www.skillsdirectory.com/skills/bordenet-feature-development)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
Files
skill.md
---
name: feature-development
source: superpowers-plus
triggers: ["start feature", "new feature", "feature development workflow",
           "build a feature", "implement feature", "feature from scratch",
           "full development workflow", "code change", "make changes",
           "fix this", "add this", "modify code", "refactor this",
           "write code", "update the code"]
anti_triggers: ["fix bug", "debug", "small change", "quick fix", "update docs"]
description: "DEFAULT WORKFLOW for ANY code change. Orchestrates the full rigorous development lifecycle: brainstorming → think-twice → debate → progressive-harsh-review → plan-and-execute → progressive-harsh-review → commit. This fires AUTOMATICALLY for code changes unless the user explicitly opts out. Sequences existing skills so no phase is skipped."
summary: "Default for ALL code changes. Opt out only if user says 'skip the full workflow' or 'just do it quickly'. Brainstorm → think-twice → design → review → plan → implement → review → ship."
coordination:
  group: engineering
  order: 1
  requires: []
  enables: ["brainstorming", "think-twice", "debate",
            "progressive-code-review-gate", "plan-and-execute",
            "requirements-validation", "todo-management",
            "output-verification", "verification-before-completion"]
  escalates_to: ["thinking-orchestrator"]
  internal: false
composition:
  consumes: [feature-request, user-story, requirement, code-change, bug-fix, refactor]
  produces: [implementation, tested-feature, completed-feature]
  capabilities: [orchestrates-workflow, sequences-skills]
  priority: 5
---

# Feature Development

> **Wrong skill?** Bug fix → `systematic-debugging`. Design comparison → `debate`. Pre-commit checks → `pre-commit-gate`. Workflow
>
> **Purpose:** Orchestrate the full rigorous development lifecycle so no phase is skipped.
> **Pattern:** This skill SEQUENCES existing skills — it does not replace them.
> **Scope:** This is the DEFAULT workflow for ANY code change. Opt out ONLY if the user explicitly says to skip it.

**Announce at start:** "I'm using the **feature-development** skill to orchestrate this workflow."

## When to Use

- Implementing a new feature end-to-end (design → code → test → ship)
- When the task spans multiple files and requires coordination
- Multi-step implementation that needs a structured workflow

## When to Use — DEFAULT for Code Changes

- **Any code change** — features, bug fixes, refactors, skill edits, config changes
- This fires AUTOMATICALLY. You do not need the user to say "use the full workflow."
- If you are about to write, edit, or generate code, this skill applies.

## Opt-Out Conditions

- User explicitly says "skip the workflow", "just do it", "quick fix only"
- Pure documentation-only changes (no code files touched)
- Reading/exploring code with no intent to change it

## The Workflow

```bash
Phase 1: BRAINSTORM → Phase 2: FRESH PERSPECTIVE → Phase 3: DESIGN →
Phase 4: HARSH REVIEW → Phase 5: PLAN & EXECUTE → Phase 6: HARSH REVIEW →
Phase 7: SHIP
```

Each phase has an **exit gate** — you cannot proceed until the gate passes.

### Phase 1: Brainstorming

**Invoke:** `brainstorming`

1. Explore the user's intent, requirements, and constraints
2. Gather codebase context — read existing code, understand patterns
3. Surface assumptions, edge cases, and scope boundaries
4. **Exit gate:** Clear understanding of what needs to change and why.

### Phase 2: Fresh Perspective

**Invoke:** `think-twice`

1. Dispatch a sub-agent with zero prior context to review the problem
2. The sub-agent should identify gaps, blind spots, and alternative framings
3. Integrate fresh insights back into the plan
4. **Exit gate:** Fresh perspective reviewed, no unaddressed blind spots.

### Phase 3: Design

**Invoke:** `debate`

1. Generate ≥3 genuinely distinct design options
2. Build comparison matrix (≤5 criteria)
3. Evaluate options against requirements from Phase 1
4. Select winning design with documented rationale
5. **Exit gate:** Design selected, trade-offs documented, edge cases addressed.

### Phase 4: Harsh Review (Design)

**Invoke:** `progressive-code-review-gate`

1. Red-team the selected design via hostile sub-agent reviewer
2. Reviewer should find issues the designer is blind to
3. Fix all BLOCKER and MAJOR findings before proceeding
4. **Exit gate:** All BLOCKER/MAJOR findings resolved. MINOR findings tracked.

### Phase 5: Plan & Execute

**Invoke:** `plan-and-execute`

1. Break the design into ordered implementation phases with success criteria
2. **Enroll each phase as a TODO** via `todo-management` / `todo-crud.sh`, then mirror to MCP tasks
3. Execute each phase, running tests after each
4. For each phase: implement → test → self-review via `adversarial-search`
5. Run `output-verification` after generating any artifact
6. **Exit gate:** All phases complete, all tests pass, no regressions.

### Phase 6: Harsh Review (Implementation)

**Invoke:** `progressive-code-review-gate`

1. Red-team the FULL implementation via hostile sub-agent reviewer
2. Reviewer reads ALL changed files, runs quality gates, checks for regressions
3. Fix all BLOCKER and MAJOR findings
4. Re-review after fixes (minimum 2 review rounds total)
5. **Exit gate:** Two review rounds passed, all findings resolved.

### Phase 7: Ship

1. Run `output-verification` — read back all generated files
2. Run `verification-before-completion` — evidence before assertions
3. Run all validation tools (harsh-review, trigger-validator, tests)
4. Commit, push, create PR, merge (with user approval at each step)
5. **Exit gate:** All checks pass, PR merged, synced to all remotes.

## Phase Skip Prevention

**You MUST NOT skip phases.** "I already know the approach" → think-twice catches blind spots. "Only one design option" → debate requires ≥3. "Too small" → the 2026-03-27 incident was "just a small script."

**Opt-out is user-initiated ONLY.** The agent never decides to skip.

## Integration Map

| Phase | Skill Invoked | Purpose |
|-------|--------------|---------|
| Brainstorm | `brainstorming` | Explore intent, gather context, surface assumptions |
| Fresh Perspective | `think-twice` | Sub-agent catches blind spots |
| Design | `debate` | ≥3 options, comparison, selection |
| Harsh Review (Design) | `progressive-code-review-gate` | Red-team the design |
| Plan & Execute | `plan-and-execute` + `adversarial-search` | Structured implementation with self-review |
| Harsh Review (Impl) | `progressive-code-review-gate` | Red-team the implementation (min 2 rounds) |
| Ship | `output-verification`, `verification-before-completion` | Inspect output, verify completion, merge |

## Incident History

**2026-03-27:** Agent skipped brainstorming, think-twice, and harsh review. Confabulated summary without reading output. Required 2 hostile review rounds. Every issue would have been caught by the full workflow.

## Example

```bash
# Pre-flight: verify branch, tests pass, no uncommitted changes
git status --short && npm test 2>&1 | tail -5
# Post-implementation: run full test suite + lint
npm run lint && npm test
```

## Failure Modes

| Failure | Recovery |
|---------|----------|
| Phase skipping (most common) | Every phase has an exit gate. No phase can be skipped without user opt-out. |
| Rushing through brainstorming in <2 exchanges | Brainstorming should explore ≥3 approaches with real trade-offs |
| Skipping harsh review after implementation | Phase 6 is mandatory. Fixes from review need their own review. |
| Not persisting phase state via TODO | Each phase must be enrolled as a TODO before starting |

## Companion Skills

- **brainstorming**: Generating options and alternatives
- **debate**: Evaluating design alternatives
- **plan-and-execute**: Structured planning workflow
- **progressive-code-review-gate**: Code review enforcement
- **test-driven-development**: TDD workflow
- **subagent-driven-development**: Multi-agent development
- **autonomous-chain-controller**: Full workflow orchestration

Attribution

bordenetbordenet
View sourceSee grades on GitHubMore from bordenet →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Terse caveman voice: answer first, fluff gone, every technical fact kept. Use for /caveman, "caveman mode", "talk like caveman", "be brief", "less tokens". Stays on until "stop caveman" or "normal mode".

1100021 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

698621 votes

Writing Skills

Create and manage Claude Code skills in HASH repository following Anthropic best practices. Use when creating new skills, modifying skill-rules.json, understanding trigger patterns, working with hooks, debugging skill activation, or implementing progressive disclosure. Covers skill structure, YAML frontmatter, trigger types (keywords, intent patterns), UserPromptSubmit hook, and the 500-line rule. Includes validation and debugging with SKILL_DEBUG. Examples include rust-error-stack, cargo-dep...

3931 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3421 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Amp, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Grok Build, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

741 votes
View all in ai-agents →