parcadeiCommunity·144 Agent Skills Included

Continuous-Claude

Self-improving Claude Code setup with memory, agents, and workflows

Turns Claude Code into a persistent, learning development environment with 109 skills, 32 agents, and 30 hooks. Eliminates lost context between sessions, wasted tokens on file reads, and repeated manual workflows. Adds memory recall, session handoffs, code analysis, testing, security audits, and math tools triggered by plain English.

3,934|299

↗

npx skills add parcadei/Continuous-Claude --all -g -y

All Skills in This Repository (144)

Pure Emerald Level Indicators

📦 In Repo

parcadei

trace-claude-code

Trace Claude Code sessions, turns, and tool calls to Braintrust.

multi-tool-pipeline

Chain multiple MCP tools into a single executable script pipeline.

braintrust-analyze

Analyze Braintrust tracing data from Claude Code sessions for patterns and insights.

onboard

Analyze a brownfield codebase and create a continuity ledger.

firecrawl-scrape

Scrape web pages and extract content via Firecrawl MCP into Markdown, HTML, or plain text.

perplexity-search

Conduct AI web searches with mode selection and model variants.

ast-grep-find

Search and refactor code by AST structural pattern matching across multi-language codebases.

github-search

Search GitHub code, repositories, issues, and pull requests.

repoprompt

Render code trees, search with context, and extract signatures via RepoPrompt CLI.

test-driven-development

Enforce test-first development using Red-Green-Refactor with failing tests before production code.

research

Documents codebase structure by spawning parallel sub-agents and synthesizing findings into research reports.

commit

Creates git commits with user approval and no Claude attribution.

Community

Intermediate

Frequently Asked Questions

FAQPage Schema

How to install Continuous-Claude?▼

Run `npx skills add parcadei/Continuous-Claude --all -g -y` in your terminal to install all skills in this suite globally.

What problem does Continuous Claude solve?▼

It stops Claude Code from losing context when sessions compact, by saving learnings and handoffs to a database. Each session starts smarter because past decisions and fixes are recalled automatically.

Do I need to memorize slash commands?▼

No. You describe what you want in plain English, and a hook suggests the right skills and agents automatically. Power users can still call commands like /fix or /build directly.

What are the system requirements?▼

You need Python 3.11+, the uv package manager, Docker for the PostgreSQL database, and the Claude Code CLI installed.

How does it save tokens on code exploration?▼

The TLDR analysis layer summarizes code structure, call graphs, and data flow instead of reading entire files, cutting token usage by roughly 95% on large codebases.