ZaaliMohamed123/evolving-claude
A self-improving global Claude Code setup: 10 subagents, on-demand skills, safety hooks, three-tier memory, and a config feedback loop.
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
Copy this rig
# review before running: this installs third-party code $ npx degit ZaaliMohamed123/evolving-claude/agents ./rig-evolving-claude/agents $ npx degit ZaaliMohamed123/evolving-claude/skills ./rig-evolving-claude/skills $ npx degit ZaaliMohamed123/evolving-claude/hooks ./rig-evolving-claude/hooks
MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ npx degit ZaaliMohamed123/evolving-claude/skills/commit .claude/skills/commit $ npx degit ZaaliMohamed123/evolving-claude/skills/fix-issue .claude/skills/fix-issue $ npx degit ZaaliMohamed123/evolving-claude/skills/init-project .claude/skills/init-project $ npx degit ZaaliMohamed123/evolving-claude/skills/onboard-codebase .claude/skills/onboard-codebase $ npx degit ZaaliMohamed123/evolving-claude/skills/open-pr .claude/skills/open-pr $ npx degit ZaaliMohamed123/evolving-claude/skills/review-diff .claude/skills/review-diff $ npx degit ZaaliMohamed123/evolving-claude/skills/tdd .claude/skills/tdd $ npx degit ZaaliMohamed123/evolving-claude/skills/update-config .claude/skills/update-config
$ curl -fsSL --create-dirs -o .claude/agents/code-reviewer.md https://raw.githubusercontent.com/ZaaliMohamed123/evolving-claude/main/agents/code-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/config-manager.md https://raw.githubusercontent.com/ZaaliMohamed123/evolving-claude/main/agents/config-manager.md $ curl -fsSL --create-dirs -o .claude/agents/debugger.md https://raw.githubusercontent.com/ZaaliMohamed123/evolving-claude/main/agents/debugger.md $ curl -fsSL --create-dirs -o .claude/agents/deep-researcher.md https://raw.githubusercontent.com/ZaaliMohamed123/evolving-claude/main/agents/deep-researcher.md $ curl -fsSL --create-dirs -o .claude/agents/docs-writer.md https://raw.githubusercontent.com/ZaaliMohamed123/evolving-claude/main/agents/docs-writer.md $ curl -fsSL --create-dirs -o .claude/agents/explore.md https://raw.githubusercontent.com/ZaaliMohamed123/evolving-claude/main/agents/explore.md $ curl -fsSL --create-dirs -o .claude/agents/implementer.md https://raw.githubusercontent.com/ZaaliMohamed123/evolving-claude/main/agents/implementer.md $ curl -fsSL --create-dirs -o .claude/agents/planner.md https://raw.githubusercontent.com/ZaaliMohamed123/evolving-claude/main/agents/planner.md $ curl -fsSL --create-dirs -o .claude/agents/security-reviewer.md https://raw.githubusercontent.com/ZaaliMohamed123/evolving-claude/main/agents/security-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/test-engineer.md https://raw.githubusercontent.com/ZaaliMohamed123/evolving-claude/main/agents/test-engineer.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 7,204 rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(rm -rf *)",
"Read(**/*.pem)",
"Bash(rm -rf /)",
"Bash(git push --force:*)",
"Bash(sudo *)",
"Read(.env)",
"Bash(rm -rf /*)",
"Read(./.env.*)",
"Read(~/.aws/**)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(sudo:*)",
"Bash(git reset --hard*)",
"Bash(git reset --hard:*)",
"Read(.env.*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(rm:*)",
"Bash(git rebase *)",
"Bash(wget *)",
"Bash(npm publish:*)",
"Bash(git commit *)",
"Bash(gh pr merge *)"
]
}
} Skills (8)
Subagents (10)
| code-reviewer model: opus | Reviews a diff or a piece of code for correctness, clarity, and consistency with the codebase. Use PROACTIVELY after any non-trivial change, in a fresh context, before considering work done. Reports g |
| config-manager model: opus | Owns the global Claude Code setup and its self-improvement loop. Use via the /update-config skill (or directly) to review the config-review-queue, propose and apply improvements to global agents/skill |
| debugger model: opus | Reproduces and root-causes bugs. Use when something fails, errors, or behaves wrong and the cause isn't obvious. Forms hypotheses, gathers evidence, isolates the root cause, and proposes (or applies) |
| deep-researcher model: opus | Thorough multi-source investigation across the codebase, docs, and external references. Use when a question needs synthesis from many places (architecture decisions, "how did this API evolve", cross-c |
| docs-writer model: haiku | Writes and updates documentation — READMEs, docstrings, API docs, ADRs, changelogs. Use when human-facing documentation needs creating or updating. Writes to docs/ and code comments, never to .claude/ |
| explore model: haiku | Fast, cheap codebase search and mapping. Use PROACTIVELY for any open-ended search, "where is X", "how does Y work", or mapping structure — anything that would read many files. Returns concise finding |
| implementer model: sonnet | Focused coding agent that executes a defined change or plan. Use when the approach is already clear (from a plan or a precise request) and you want the implementation done in an isolated context. Writ |
| planner model: opus | Authors detailed implementation plans and specs before code is written. Use for any non-trivial or multi-file feature. Explores read-only, then produces a concrete, self-contained plan (files to touch |
| security-reviewer model: opus | Security-focused review of code or a diff. Use before shipping anything that handles auth, user input, secrets, file paths, or external requests. Read-only. Reports concrete vulnerabilities with line |
| test-engineer model: sonnet | Writes and runs tests. Use for adding test coverage, building a failing test to reproduce a bug (TDD), or hardening a change with edge-case tests. Iterates until tests pass or a real defect is proven. |
Similar rigs
Yassinekrn/claude-setup
Personal Claude Code setup — orchestrator/subagent delegation policy, custom agents, and a colored PowerShell statusline (git branch, model+effort, tokens, cost, rate limits).
Orchestrator 1.2k tok ·
justxor/claude-code-pro-course
Claude Code PRO — полный курс и гайд на русском: CLAUDE.md, settings, skills, subagents, hooks, MCP, плагины, CI. 8 лабораторных, библиотека промптов и готовый starter-kit. Здесь бесплатные курсы по chatgpt -
Fort Knox 27.4k tok ·
Joan266/claude-dotfiles
—
Pragmatist 831 tok ·
tylerjkells/claude-dotfiles
—
Pragmatist 579 tok ·