Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
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
  • 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

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

Back to skills

Direct Implementation

ASecurity

Rapid single-turn direct implementation workflow for unambiguous, low-risk,

3 stars
0 votes
0 copies
0 views
Added 9/20/2026
code-qualitytestingrefactoringgitapi

Works with

cursorapi

Security Analysis

A100/100

Scanned 9/20/2026

Install to Claude Code

$npx -y skills add ruskicoder/system-prompts --skill direct-implementation --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Direct Implementation?

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

Security grade badge for Direct Implementation
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/ruskicoder-direct-implementation-bfc55035/badge)](https://www.skillsdirectory.com/skills/ruskicoder-direct-implementation-bfc55035)

More formats (shields.io, HTML) on the badges page.

Download Zip
Files
SKILL.md
---
name: direct-implementation
description: Rapid single-turn direct implementation workflow for unambiguous, low-risk,
  single-file code changes and bug fixes with known root causes. Enforces pre-action
  checks and post-edit sanity verification.
argument-hint: <change description or target file>
---

<!-- Generated from workflows/direct-implementation.md by tools/generate_integrations.py. Edit the source file, not this one. This is an execution WORKFLOW packaged as an Agent Skill so it is discoverable and directly invocable ("/direct-implementation") in every compatible tool. -->

# Workflow: Direct Implementation

## When to Use
| Criteria | Match |
|----------|-------|
| Intention | simple_change, bug_fix (known cause) |
| Complexity | single file, known solution, low risk |
| Power Mode | Eco or Balanced |
| Clarity | requirements are precise and unambiguous |

## Required Skills
- file-operations
- code-execution
- code-quality-testing
- communication-tone
- safety-profiles (default)

## Flow

### Step 0: Thinking Stage (Pre-Action)
Before any action, apply the Universal Pre-Action Protocol from AGENT.md §B:
- [ ] Do I understand the user's actual intent? Re-evaluate on every message *(Factory)*
- [ ] Have I read the relevant files? NEVER edit code you haven't read
- [ ] Have I verified my assumptions? Search/grep/query to confirm before acting
- [ ] Do I have enough context, or am I guessing?
- [ ] If unsure: search first, ask user second

### Step 1: Understand
- Read the request once
- If unclear: ask ONE clarifying question, then proceed
- If clear: acknowledge and move to execute

### Step 2: Locate
- Search/grep for relevant files
- Read the file(s) you need to modify
- Do NOT explore unrelated code

### Step 3: Edit
- Make the minimal change to satisfy the request
- Use SEARCH/REPLACE for targeted edits
- Combine all changes into minimal tool calls
- Verify: all imports, deps, and syntax correct

### Step 4: Verify
- Run linter on changed files
- Run relevant tests
- Fix any introduced errors (max 3 cycles)

### Step 4b: Sanity Check (Pre-Presentation)
Before presenting, run the Sanity Check from AGENT.md §B Stage 2:
- [ ] Is this the MINIMAL change needed? No scope creep?
- [ ] Does the code have all necessary imports and dependencies?
- [ ] Will it compile/build immediately? *(Windsurf)*
- [ ] Have I checked for linter errors? *(Cursor: max 3 fix cycles)*
- [ ] Did I verify the change fixes the actual problem?

If any check fails: STOP, fix, then proceed.

### Step 5: Present
- Brief summary: what changed, why, and (if applicable) how to verify
- Don't elaborate beyond what was asked

## Token Budget
- **Typical**: 2K-10K tokens
- **Max recommended**: 20K tokens
- **Warning threshold**: 50K tokens → you're probably in the wrong workflow

## GitNexus Integration
- Optional: `query({search_query})` for quick symbol lookup if unsure of location
- Skip for known locations

## Hallucination Watchpoints
- Overcomplicating a simple change (adding features not requested)
- Editing files without reading them first
- Making changes beyond the scope of the request
- Refactoring unrelated code "while you're in there"

## Example
```
User: "Change the port from 3000 to 8080 in server.js"

1. grep for "3000" in server.js → found at line 12
2. read server.js → understand context
3. edit: port 3000 → port 8080
4. run linter → clean
5. "Changed port from 3000 to 8080 in server.js:12"
```

Attribution

ruskicoderruskicoder
View sourceMore from ruskicoder →
SSkills DirectorySkills Directory

Your tool, in front of Claude Code builders.

3 founder slots · $299/mo · GSC-verified traffic · sponsors can never buy grades.

See placements

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

Your tool, in front of Claude Code builders.

3 founder slots · $299/mo · GSC-verified traffic · sponsors can never buy grades.

See placements

Related Skills

Caveman Review

Ultra-compressed code review comments. Cuts noise from PR feedback while preserving the actionable signal. Each comment is one line: location, problem, fix. Use when user says "review this PR", "code review", "review the diff", "/review", or invokes /caveman-review. Auto-triggers when reviewing pull requests.

1023331 votes

Caveman Commit

Ultra-compressed commit message generator. Cuts noise from commit messages while preserving intent and reasoning. Conventional Commits format. Subject ≤50 chars, body only when "why" isn't obvious. Use when user says "write a commit", "commit message", "generate commit", "/commit", or invokes /caveman-commit. Auto-triggers when staging changes.

1023331 votes

Verification Loop

一个全面的 Claude Code 会话验证系统。

2456590 votes

Springboot Verification

Verification loop for Spring Boot projects: build, static analysis, tests with coverage, security scans, and diff review before release or PR.

2456590 votes

Django Verification

Verification loop for Django projects: migrations, linting, tests with coverage, security scans, and deployment readiness checks before release or PR.

2456590 votes
View all in code-quality →