Menu
Store universal coding patterns into vector database. Auto-invokes after difficult tasks with broadly-applicable lessons. Trigger with "--store" or when user expresses frustration (strong learning signals).
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Headless browser automation CLI for AI agents. Use this skill when you need to automate web interactions, scrape web pages, test web applications, or perform any browser-based tasks. Triggers include "browse", "automate browser", "scrape website", "test web app", "fill form", "take screenshot", or any web automation request. Provides accessibility-first element targeting with semantic locators.
Toolkit for testing Android applications using Appium with Espresso driver. Supports automated UI testing, gesture interactions, element discovery, and screenshot capture. Use when testing native Android apps (.apk files), verifying mobile UI functionality, automating gestures (swipe, scroll, tap), or running blackbox tests on Android devices/emulators.
This skill should be used when users want to create, configure, or debug Claude Code hooks. Hooks are shell commands that execute at various points in Claude Code's lifecycle (PreToolUse, PostToolUse, Stop, SessionStart, etc.). Use this skill when users ask to create custom automation, file protection, code formatting hooks, notifications, or any lifecycle-based triggers for Claude Code.
Retrieve universal coding patterns from vector database. Auto-invokes before complex tasks or when user says "--recall". Searches relevant role collections based on task context.
Create custom slash commands for Claude Code. Use this skill when the user asks to "create a command", "add a slash command", "make a custom command", "write a command for...", or needs help with command frontmatter, arguments ($ARGUMENTS, $1, $2), file references (@syntax), bash execution in commands, or organizing commands in project/personal/plugin locations.
Find reputable GitHub repositories with quality filters (stars, recency, license). Use this skill when the user needs to find code samples, reference implementations, libraries, or templates on GitHub. Triggers on requests like "find GitHub repos for X", "search GitHub for Y examples", "find a good library for Z", or any request to discover quality open-source code. Requires gh CLI (version 2.14+).
| name | coder-memory-store |
| description | Store universal coding patterns into vector database. Auto-invokes after difficult tasks with broadly-applicable lessons. Trigger with "--store" or when user expresses frustration (strong learning signals). |
The memory-only agent has ZERO access to Read/Write/Edit/Glob/Bash - it can ONLY use MCP memory tools. This prevents file reading pollution by design.
Store 0-1 insights per task (rarely 2-3).Store when: Non-obvious bugs, hard-won failure lessons, universal patterns across projects, user frustration signals.
Skip when: Standard practices, project-specific config, routine fixes, vague insights.
**Step 1: Format memory** ``` **Title:** [Concise title] **Description:** [2-3 sentence summary - CRITICAL for search]Content: [What happened, what was tried, what worked/failed, key lesson]
Tags: #role #topic #success|#failure
**Step 2: Extract metadata**
Parse the formatted text to extract:
- `title`: Plain text without markdown (from "**Title:**")
- `description`: Plain text without markdown (from "**Description:**")
- `tags`: Array of tags (e.g., ["#backend", "#jwt", "#auth"])
**Step 3: Detect role**
Map task context to role:
- backend: api, endpoint, database, server, auth
- frontend: react, vue, component, ui, css
- devops: deploy, docker, kubernetes, ci/cd
- ai: model, training, embedding, llm
- security: vulnerability, encryption, jwt
- mobile: ios, android, flutter, swift
- scrum-master: agile, sprint, standup, retrospective, planning
Default to "universal" if unclear.
**Step 4: Search for duplicates**
Use `search_memory` with full formatted text as query, `roles=["detected_role", "universal"]`, `limit=10`.
**Step 5: Decide action**
- Near-identical exists → MERGE (combine, delete old)
- Related exists → UPDATE (enhance existing)
- Pattern emerges from 2+ episodic → GENERALIZE to semantic
- Different topic → CREATE new
**Step 6: Store**
Use `store_memory` with:
- `document`: Full formatted text (Title + Description + Content + Tags)
- `metadata`:
```json
{
"memory_type": "episodic|procedural|semantic",
"role": "backend|frontend|devops|ai|...",
"title": "Plain text title",
"description": "2-3 line summary",
"tags": ["#tag1", "#tag2"],
"confidence": "high|medium|low",
"frequency": 1
}
<role_mapping> Available roles:
Default to "universal" if unclear. </role_mapping>
<frustration_signals> User frustration = critical learning moment. When detected (profanity, "this is ridiculous", emotional language): Store as episodic with #failure #strong-signal tags. </frustration_signals>