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
  • 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

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Context Preservation

ASecurity

Generate structured context summaries for session continuity across compaction

192 stars
0 votes
0 copies
1 views
Added 2/8/2026
developmentgo

Security Analysis

A100/100

Scanned 2/12/2026

Install to Claude Code

$npx -y skills add juanandresgs/claude-system --skill context-preservation --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Context Preservation?

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

Security grade badge for Context Preservation
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/juanandresgs-context-preservation/badge)](https://www.skillsdirectory.com/skills/juanandresgs-context-preservation)

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

Download with Pro
Files
SKILL.md
---
name: context-preservation
description: Generate structured context summaries for session continuity across compaction
---

# Context Preservation Skill

Generate structured context summaries when compaction is imminent or requested, ensuring session continuity without information loss.

## Output Format (STRICT - NO DEVIATIONS)

```markdown
### 1. Current Objective & Status
- **Goal**: [Robust, multi-sentence description of what we are building. Include critical details, context, and the "Definition of Done". Do not summarize into a single vague line.]
- **Status**: [Completed | In Progress | Blocked]
- **Immediate Next Step**: [The very next command or code edit required. Be specific.]

### 2. Active Context
- **Active Files**:
  - `[Absolute Path]`
  - `[Absolute Path]`
- **Recent Changes**: [Specific descriptions of what *just* changed in the code, referencing specific functions or logic.]
- **Variables/State**: [Key variable names, data structures, or temporary states currently in focus.]

### 3. Constraints & Decisions (CRITICAL)
- **Preferences**: [User preferences stated in this session, e.g., "no external deps", "use snake_case"]
- **Discarded Approaches**: [What did we try that failed? Do not repeat these mistakes.]
- **Architectural Rules**: [Patterns we agreed on, e.g., "Service Layer pattern", "DTOs required"]

### 4. Continuity Handoff
- "When resuming, the first thing to do is..."
```

## Content Extraction

### Section 1: Current Objective & Status
- Extract goal from user's initial request, explicit goal statements, Definition of Done indicators
- Status: `Completed` (all objectives achieved), `In Progress` (work ongoing), `Blocked` (explicit blocker)
- Next step: last in-progress item, last "next I will..." statement, or first pending objective

### Section 2: Active Context
- Files touched via Read/Write/Edit operations, prioritized by recency
- **ALWAYS use absolute paths** (e.g., `/Users/turla/Code/project/src/auth.ts`)
- Summarize function additions, modifications, deletions with specific names and line refs
- Key variables, data structures, environment variables in focus

### Section 3: Constraints & Decisions
- Preferences: "don't use...", "prefer...", "always...", "never..."
- Discarded approaches: failed attempts with reasons
- Architectural rules: pattern agreements, structure decisions, dependency choices

### Section 4: Continuity Handoff
- Always begin with "When resuming, the first thing to do is..."
- Must be specific enough for a fresh Claude instance to immediately continue

## Anti-Patterns (NEVER DO)

### Goal Description

**WRONG**:
- "Building a feature"
- "Working on authentication"

**RIGHT**:
- "Building a user authentication system with OAuth2 PKCE flow for the mobile app. The system must support Google and Apple SSO, store refresh tokens securely in the device keychain. Definition of Done: user can complete full login flow, tokens persist across app restarts, logout clears all tokens."

### Active Files

**WRONG**:
- `file.ts`
- `./src/auth.ts`

**RIGHT**:
- `/Users/turla/Code/myproject/src/auth/oauth-handler.ts`

### Recent Changes

**WRONG**:
- "Updated the file"
- "Fixed some issues"

**RIGHT**:
- "Added `validateTokenExpiry()` function to `/Users/turla/Code/myproject/src/auth/token-storage.ts` (lines 42-67) that checks JWT expiration. Modified `refreshToken()` to call this before attempting refresh."

### Immediate Next Step

**WRONG**:
- "Continue working"
- "Finish the implementation"

**RIGHT**:
- "Run `npm test -- --testPathPattern=oauth-handler` to verify the new `validateTokenExpiry()` function works with expired tokens"

## Validation Checklist

### Section 1
- [ ] Goal is 2+ sentences with Definition of Done
- [ ] Status is exactly: Completed, In Progress, or Blocked
- [ ] Next step is a specific command or edit

### Section 2
- [ ] All file paths are absolute (start with `/`)
- [ ] Recent changes reference specific function/method names

### Section 3
- [ ] Discarded approaches explain WHY they were discarded
- [ ] No empty sections (use "None stated this session" if applicable)

### Section 4
- [ ] Starts with "When resuming, the first thing to do is..."
- [ ] Contains a specific, actionable instruction

Attribution

juanandresgsjuanandresgs
View sourceMore from juanandresgs →
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

Browser Extension Developer

Use this skill when developing or maintaining browser extension code in the `browser/` directory, including Chrome/Firefox/Edge compatibility, content scripts, background scripts, or i18n updates.

284072 votes

Seo Optimizer

SEO optimization with keyword analysis, readability assessment, technical validation, content quality. Use for search rankings, blog posts, content audits, or encountering keyword density, readability scores, meta tags, schema markup errors.

2192 votes

Google Official Seo Guide

Official Google SEO guide covering search optimization, best practices, Search Console, crawling, indexing, and improving website search visibility based on official Google documentation

1862 votes

Tanstack Start

Build a full-stack TanStack Start app on Cloudflare Workers from scratch — SSR, file-based routing, server functions, D1+Drizzle, better-auth, Tailwind v4+shadcn/ui. Use whenever the user mentions TanStack Start, asks to scaffold a full-stack Cloudflare app with SSR, wants an SSR dashboard, or asks for a React 19 + Cloudflare Workers app with file-based routing and server functions — even if they don't name TanStack Start specifically. No template repo — Claude generates every file fresh per ...

9881 votes

Pentest

PTES-aligned adversarial security audit for backend, frontend, and mobile applications. Produces a CVSS-scored Hacker Report with verified PoCs and phased remediation.

5491 votes
View all in development →