mhxie/Atelier
AI-native personal operating system
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 mhxie/Atelier/.claude ./rig-atelier # inspect, then merge into .claude/ 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 mhxie/Atelier/.agents/skills/atelier .claude/skills/atelier $ npx degit mhxie/Atelier/.agents/skills/autoevo-review .claude/skills/autoevo-review $ npx degit mhxie/Atelier/.agents/skills/civ .claude/skills/civ $ npx degit mhxie/Atelier/.agents/skills/curate .claude/skills/curate $ npx degit mhxie/Atelier/.agents/skills/daily-reflection .claude/skills/daily-reflection $ npx degit mhxie/Atelier/.agents/skills/decision .claude/skills/decision $ npx degit mhxie/Atelier/.agents/skills/digest .claude/skills/digest $ npx degit mhxie/Atelier/.agents/skills/dine .claude/skills/dine $ npx degit mhxie/Atelier/.agents/skills/hi .claude/skills/hi $ npx degit mhxie/Atelier/.agents/skills/introspect .claude/skills/introspect $ npx degit mhxie/Atelier/.agents/skills/lint .claude/skills/lint $ npx degit mhxie/Atelier/.agents/skills/prm .claude/skills/prm $ npx degit mhxie/Atelier/.agents/skills/promote .claude/skills/promote $ npx degit mhxie/Atelier/.agents/skills/push .claude/skills/push $ npx degit mhxie/Atelier/.agents/skills/read .claude/skills/read $ npx degit mhxie/Atelier/.agents/skills/reflect .claude/skills/reflect $ npx degit mhxie/Atelier/.agents/skills/review .claude/skills/review $ npx degit mhxie/Atelier/.agents/skills/triage .claude/skills/triage $ npx degit mhxie/Atelier/.agents/skills/weekly .claude/skills/weekly $ npx degit mhxie/Atelier/.claude/skills/capture .claude/skills/capture $ npx degit mhxie/Atelier/.claude/skills/reading .claude/skills/reading
$ curl -fsSL --create-dirs -o .claude/agents/challenger.md https://raw.githubusercontent.com/mhxie/Atelier/main/.claude/agents/challenger.md $ curl -fsSL --create-dirs -o .claude/agents/curator.md https://raw.githubusercontent.com/mhxie/Atelier/main/.claude/agents/curator.md $ curl -fsSL --create-dirs -o .claude/agents/forgetter.md https://raw.githubusercontent.com/mhxie/Atelier/main/.claude/agents/forgetter.md $ curl -fsSL --create-dirs -o .claude/agents/librarian.md https://raw.githubusercontent.com/mhxie/Atelier/main/.claude/agents/librarian.md $ curl -fsSL --create-dirs -o .claude/agents/meeting.md https://raw.githubusercontent.com/mhxie/Atelier/main/.claude/agents/meeting.md $ curl -fsSL --create-dirs -o .claude/agents/precedent-judge.md https://raw.githubusercontent.com/mhxie/Atelier/main/.claude/agents/precedent-judge.md $ curl -fsSL --create-dirs -o .claude/agents/privacy-reviewer.md https://raw.githubusercontent.com/mhxie/Atelier/main/.claude/agents/privacy-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/reader.md https://raw.githubusercontent.com/mhxie/Atelier/main/.claude/agents/reader.md $ curl -fsSL --create-dirs -o .claude/agents/researcher.md https://raw.githubusercontent.com/mhxie/Atelier/main/.claude/agents/researcher.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer.md https://raw.githubusercontent.com/mhxie/Atelier/main/.claude/agents/reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/scholar.md https://raw.githubusercontent.com/mhxie/Atelier/main/.claude/agents/scholar.md $ curl -fsSL --create-dirs -o .claude/agents/scout.md https://raw.githubusercontent.com/mhxie/Atelier/main/.claude/agents/scout.md $ curl -fsSL --create-dirs -o .claude/agents/scribe.md https://raw.githubusercontent.com/mhxie/Atelier/main/.claude/agents/scribe.md $ curl -fsSL --create-dirs -o .claude/agents/synthesizer.md https://raw.githubusercontent.com/mhxie/Atelier/main/.claude/agents/synthesizer.md $ curl -fsSL --create-dirs -o .claude/agents/thinker.md https://raw.githubusercontent.com/mhxie/Atelier/main/.claude/agents/thinker.md $ curl -fsSL --create-dirs -o .claude/agents/challenger.md https://raw.githubusercontent.com/mhxie/Atelier/main/agents/challenger.md $ curl -fsSL --create-dirs -o .claude/agents/curator.md https://raw.githubusercontent.com/mhxie/Atelier/main/agents/curator.md $ curl -fsSL --create-dirs -o .claude/agents/forgetter.md https://raw.githubusercontent.com/mhxie/Atelier/main/agents/forgetter.md $ curl -fsSL --create-dirs -o .claude/agents/librarian.md https://raw.githubusercontent.com/mhxie/Atelier/main/agents/librarian.md $ curl -fsSL --create-dirs -o .claude/agents/meeting.md https://raw.githubusercontent.com/mhxie/Atelier/main/agents/meeting.md $ curl -fsSL --create-dirs -o .claude/agents/precedent-judge.md https://raw.githubusercontent.com/mhxie/Atelier/main/agents/precedent-judge.md $ curl -fsSL --create-dirs -o .claude/agents/privacy-reviewer.md https://raw.githubusercontent.com/mhxie/Atelier/main/agents/privacy-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/reader.md https://raw.githubusercontent.com/mhxie/Atelier/main/agents/reader.md $ curl -fsSL --create-dirs -o .claude/agents/researcher.md https://raw.githubusercontent.com/mhxie/Atelier/main/agents/researcher.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer.md https://raw.githubusercontent.com/mhxie/Atelier/main/agents/reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/scholar.md https://raw.githubusercontent.com/mhxie/Atelier/main/agents/scholar.md $ curl -fsSL --create-dirs -o .claude/agents/scout.md https://raw.githubusercontent.com/mhxie/Atelier/main/agents/scout.md $ curl -fsSL --create-dirs -o .claude/agents/scribe.md https://raw.githubusercontent.com/mhxie/Atelier/main/agents/scribe.md $ curl -fsSL --create-dirs -o .claude/agents/synthesizer.md https://raw.githubusercontent.com/mhxie/Atelier/main/agents/synthesizer.md $ curl -fsSL --create-dirs -o .claude/agents/thinker.md https://raw.githubusercontent.com/mhxie/Atelier/main/agents/thinker.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 (21)
atelierautoevo-reviewcivcuratedaily-reflectiondecisiondigestdinehiintrospectlintprmpromotepushreadreflectreviewtriageweeklycapturereading
Subagents (30)
| challenger model: opus | Presents probing questions to affirm or challenge the user's latest thoughts and feelings. Use to deepen reflection and test assumptions. |
| curator model: sonnet | Drafts note-content operations — compactions, merges, new notes, wiki entries — for the orchestrator to write under $OV/. Use when the user wants to consolidate, restructure, or create local notes. |
| forgetter model: sonnet | Active decay scanner over $OV/. Finds what no longer earns its place — redundant, time-stale, contradicted, or low-signal. Proposes; never deletes. Returns categorized findings inline for a trusted pa |
| librarian model: sonnet | Recommends readings, resources, and thinkers relevant to the user's current interests and goals. Use when the user wants learning recommendations. |
| meeting model: sonnet | Processes work meeting transcripts into structured notes with action items, decisions, and next steps. For research talks and presentations, use Reader instead. Le cercle archetype — The Stenographer. |
| precedent-judge model: sonnet | Applies the user's own past decisions to a new pending item. Answers supplied prompts as one JSON verdict each, inline for Autoevo proposals or file-backed for interactive batches, and never invents p |
| privacy-reviewer model: sonnet | Semantic privacy scanner for committed-file diffs. Catches leaks the mechanical scripts/privacy_check.py misses, including contextual identity, financial, demographic, and personal-taxonomy disclosure |
| reader model: sonnet | Reads articles and notes through requested perspectives (Critical, Structural, Practical, Dialectical), including transcript preprocessing. One reader can cover multiple lenses. For long or hard reads |
| researcher model: opus | Gathers raw context from the user's local $OV/ vault (daily notes, reflections, wiki, papers). Use when you need to pull notes, search for themes, or collect evidence before synthesis. |
| reviewer model: sonnet | Quality-checks reflection, reading, and synthesis outputs against their source evidence in Session Review mode, and wiki claim edits in Claim Review mode. Reports actual defects with severity and cove |
| scholar model: opus | Deep-cognition variant of Reader for dense theory, foundational papers, and hard texts. Same requested perspectives and workflow with stronger voices; one scholar can cover multiple lenses. Reader han |
| scout model: sonnet | Gathers external context from the web — articles, discussions, research, recent developments. The team's eyes into the outside world. |
| scribe model: haiku | Records user-dictated raw content verbatim to a named target file. Handles five operations — daily-note capture, dining-log row append, GTD entry append/toggle, people-note stub create, generic raw pa |
| synthesizer model: opus | Reads gathered context and produces structured reflections, summaries, and insights. Use after the Researcher has gathered raw material. |
| thinker model: opus | Thinks independently using structured frameworks. Offers fresh perspectives unconstrained by existing context. Use for outside views or cross-validation. |
| challenger model: opus | Presents probing questions to affirm or challenge the user's latest thoughts and feelings. Use to deepen reflection and test assumptions. |
| curator model: sonnet | Drafts note-content operations — compactions, merges, new notes, wiki entries — for the orchestrator to write under $OV/. Use when the user wants to consolidate, restructure, or create local notes. |
| forgetter model: sonnet | Active decay scanner over $OV/. Finds what no longer earns its place — redundant, time-stale, contradicted, or low-signal. Proposes; never deletes. Returns categorized findings inline for a trusted pa |
| librarian model: sonnet | Recommends readings, resources, and thinkers relevant to the user's current interests and goals. Use when the user wants learning recommendations. |
| meeting model: sonnet | Processes work meeting transcripts into structured notes with action items, decisions, and next steps. For research talks and presentations, use Reader instead. Le cercle archetype — The Stenographer. |
| precedent-judge model: sonnet | Applies the user's own past decisions to a new pending item. Answers supplied prompts as one JSON verdict each, inline for Autoevo proposals or file-backed for interactive batches, and never invents p |
| privacy-reviewer model: sonnet | Semantic privacy scanner for committed-file diffs. Catches leaks the mechanical scripts/privacy_check.py misses, including contextual identity, financial, demographic, and personal-taxonomy disclosure |
| reader model: sonnet | Reads articles and notes through requested perspectives (Critical, Structural, Practical, Dialectical), including transcript preprocessing. One reader can cover multiple lenses. For long or hard reads |
| researcher model: opus | Gathers raw context from the user's local $OV/ vault (daily notes, reflections, wiki, papers). Use when you need to pull notes, search for themes, or collect evidence before synthesis. |
| reviewer model: sonnet | Quality-checks reflection, reading, and synthesis outputs against their source evidence in Session Review mode, and wiki claim edits in Claim Review mode. Reports actual defects with severity and cove |
| scholar model: opus | Deep-cognition variant of Reader for dense theory, foundational papers, and hard texts. Same requested perspectives and workflow with stronger voices; one scholar can cover multiple lenses. Reader han |
| scout model: sonnet | Gathers external context from the web — articles, discussions, research, recent developments. The team's eyes into the outside world. |
| scribe model: haiku | Records user-dictated raw content verbatim to a named target file. Handles five operations — daily-note capture, dining-log row append, GTD entry append/toggle, people-note stub create, generic raw pa |
| synthesizer model: opus | Reads gathered context and produces structured reflections, summaries, and insights. Use after the Researcher has gathered raw material. |
| thinker model: opus | Thinks independently using structured frameworks. Offers fresh perspectives unconstrained by existing context. Use for outside views or cross-validation. |
Hooks (13)
| event | matcher | runs |
|---|---|---|
| SessionStart | * | uv run "${CLAUDE_PROJECT_DIR:-.}/scripts/cues.py" --hook --runtime claude |
| SessionStart | * | python3 "${CLAUDE_PROJECT_DIR:-.}/scripts/observability/native.py" hook claude 2>/dev/null || true |
| UserPromptSubmit | * | python3 "${CLAUDE_PROJECT_DIR:-.}/scripts/autoevo_preflight.py" --touch-lock 2>/dev/null || true |
| UserPromptSubmit | * | python3 "${CLAUDE_PROJECT_DIR:-.}/scripts/observability/native.py" hook claude 2>/dev/null || true |
| PostToolUse | * | python3 "${CLAUDE_PROJECT_DIR:-.}/scripts/autoevo_preflight.py" --touch-lock 2>/dev/null || true |
| PostToolUse | Write | [ -f ~/.claude/scripts/peer_review_gate.py ] && python3 ~/.claude/scripts/peer_review_gate.py 2>/dev/null || true |
| SessionEnd | * | python3 "${CLAUDE_PROJECT_DIR:-.}/scripts/observability/native.py" hook claude 2>/dev/null || true |
| Stop | * | python3 "${CLAUDE_PROJECT_DIR:-.}/scripts/autoevo_preflight.py" --touch-lock 2>/dev/null || true |
| Stop | * | python3 "${CLAUDE_PROJECT_DIR:-.}/scripts/observability/native.py" hook claude 2>/dev/null || true |
| SubagentStart | * | python3 "${CLAUDE_PROJECT_DIR:-.}/scripts/observability/native.py" hook claude 2>/dev/null || true |
| SubagentStop | * | python3 "${CLAUDE_PROJECT_DIR:-.}/scripts/observability/native.py" hook claude 2>/dev/null || true |
| PostCompact | * | python3 "${CLAUDE_PROJECT_DIR:-.}/scripts/observability/native.py" hook claude 2>/dev/null || true |
| StopFailure | * | python3 "${CLAUDE_PROJECT_DIR:-.}/scripts/observability/native.py" hook claude 2>/dev/null || true |
Slash commands (18)
/autoevo-review/civ/curate/daily-reflection/decision/digest/dine/hi/introspect/lint/prm/promote/push/read/reflect/review/triage/weekly
Similar rigs
ArcticFox2029/chamnan
Preserves a long-lived repository's engineering context — an architecture index, an impact map, session records, and the decisions behind them — as markdown committed beside the code, so an agent reads instead of rediscovering. Repository-l
Skill Collector 666 tok ·
tammai/ai-sdlc
Claude Code plugin for vibe coding on the AI-native SDLC playbook: intent → design → spec → plan → verify → review → ship. Runs on a Claude subscription, with opinionated stacks (Nuxt/Next + Cloudflare, Go + Postgres, Flutter, Tauri) and de
Orchestrator 3.2k tok ·
shahshrey/unclaw
Turn Claude Code into your personal AI agent. No harness required. No API costs. works with your existing Max subscription.
YOLO Cowboy 4.6k tok ·
pruthvidarji1993/ai-automation-jira
AI-assisted dev workflow: Jira ticket in, reviewed PR out. Claude Code plugin + GitHub Copilot prompts with human approval gates (plan, diff).
Pragmatist 2.0k tok ·