cdeust/ai-architect-mcp-codebase
AI Architect Codebase: cross-platform code intelligence MCP for Claude Code, Codex, Gemini, Cursor, VS Code, and Zed. Tree-sitter AST to LadybugDB graph, hybrid search, and impact analysis.
field: unknown workflow: Frontend & designPlanning & specsPR / code reviewDebuggingSecurity review runs claude in CI
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
CONTEXT TAX · EVERY TURN
up to ~7.2k tokens
Claude Code loads MCP tools on demand, so the real cost is usually lower.
Heavy · median rig: 2.3k · breakdown
Copy this rig
# review before running: this installs third-party code
$ npx degit cdeust/ai-architect-mcp-codebase/.claude ./rig-ai-architect-mcp-codebase # 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 cdeust/ai-architect-mcp-codebase/plugins/ai-architect-mcp-codebase/skills/impact-analysis .claude/skills/impact-analysis $ npx degit cdeust/ai-architect-mcp-codebase/plugins/ai-architect-mcp-codebase/skills/understand-codebase .claude/skills/understand-codebase $ npx degit cdeust/ai-architect-mcp-codebase/plugins/ai-architect-mcp-codebase/skills/validate-change-plan .claude/skills/validate-change-plan
$ curl -fsSL --create-dirs -o .claude/agents/architect.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/architect.md $ curl -fsSL --create-dirs -o .claude/agents/code-quality-reviewer.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/code-quality-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/code-reviewer.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/code-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/data-scientist.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/data-scientist.md $ curl -fsSL --create-dirs -o .claude/agents/dba.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/dba.md $ curl -fsSL --create-dirs -o .claude/agents/devops-engineer.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/devops-engineer.md $ curl -fsSL --create-dirs -o .claude/agents/documentation-accuracy-reviewer.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/documentation-accuracy-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/engineer.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/engineer.md $ curl -fsSL --create-dirs -o .claude/agents/experiment-runner.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/experiment-runner.md $ curl -fsSL --create-dirs -o .claude/agents/frontend-engineer.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/frontend-engineer.md $ curl -fsSL --create-dirs -o .claude/agents/alexander.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/alexander.md $ curl -fsSL --create-dirs -o .claude/agents/alkhwarizmi.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/alkhwarizmi.md $ curl -fsSL --create-dirs -o .claude/agents/altshuller.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/altshuller.md $ curl -fsSL --create-dirs -o .claude/agents/archimedes.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/archimedes.md $ curl -fsSL --create-dirs -o .claude/agents/arendt.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/arendt.md $ curl -fsSL --create-dirs -o .claude/agents/aristotle.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/aristotle.md $ curl -fsSL --create-dirs -o .claude/agents/bateson.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/bateson.md $ curl -fsSL --create-dirs -o .claude/agents/beer.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/beer.md $ curl -fsSL --create-dirs -o .claude/agents/borges.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/borges.md $ curl -fsSL --create-dirs -o .claude/agents/boyd.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/boyd.md $ curl -fsSL --create-dirs -o .claude/agents/braudel.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/braudel.md $ curl -fsSL --create-dirs -o .claude/agents/bruner.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/bruner.md $ curl -fsSL --create-dirs -o .claude/agents/carnot.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/carnot.md $ curl -fsSL --create-dirs -o .claude/agents/champollion.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/champollion.md $ curl -fsSL --create-dirs -o .claude/agents/coase.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/coase.md $ curl -fsSL --create-dirs -o .claude/agents/cochrane.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/cochrane.md $ curl -fsSL --create-dirs -o .claude/agents/curie.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/curie.md $ curl -fsSL --create-dirs -o .claude/agents/darwin.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/darwin.md $ curl -fsSL --create-dirs -o .claude/agents/deming.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/deming.md $ curl -fsSL --create-dirs -o .claude/agents/dijkstra.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/dijkstra.md $ curl -fsSL --create-dirs -o .claude/agents/eco.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/eco.md $ curl -fsSL --create-dirs -o .claude/agents/einstein.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/einstein.md $ curl -fsSL --create-dirs -o .claude/agents/ekman.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/ekman.md $ curl -fsSL --create-dirs -o .claude/agents/engelbart.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/engelbart.md $ curl -fsSL --create-dirs -o .claude/agents/erdos.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/erdos.md $ curl -fsSL --create-dirs -o .claude/agents/erlang.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/erlang.md $ curl -fsSL --create-dirs -o .claude/agents/euler.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/euler.md $ curl -fsSL --create-dirs -o .claude/agents/feinstein.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/feinstein.md $ curl -fsSL --create-dirs -o .claude/agents/fermi.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/fermi.md $ curl -fsSL --create-dirs -o .claude/agents/feynman.md https://raw.githubusercontent.com/cdeust/ai-architect-mcp-codebase/main/.claude/agents/genius/feynman.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 16.2k rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(./.env.*)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(git push --force:*)",
"Read(**/*.pem)",
"Read(.env)",
"Read(~/.aws/**)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Bash(rm -rf *)",
"Bash(sudo *)",
"Bash(sudo:*)",
"Bash(rm -rf /)",
"Bash(git push -f:*)",
"Read(.env.*)",
"Bash(git reset --hard*)",
"Bash(rm -rf /*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(gh pr merge *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(gh pr merge:*)",
"Bash(gh pr create *)",
"Bash(git reset --hard:*)",
"Bash(npm publish:*)",
"Bash(git rebase *)"
]
}
} MCP servers (1)
| server | source | est. tokens |
|---|---|---|
| ai-architect | local / custom | 2.5k |
Skills (3)
Subagents (79)
| architect model: opus | Software architect for module decomposition, layer boundary design, dependency analysis, and refactoring strategy |
| code-quality-reviewer model: inherit | Use this agent when you need to review code for quality, maintainability, and adherence to best practices. Examples:\n\n- After implementing a new feature or function:\n user: 'I've just written a fu |
| code-reviewer model: opus | Code reviewer specializing in Clean Architecture enforcement, SOLID violations, and architectural integrity |
| data-scientist model: opus | Data analysis and pipeline specialist — EDA, feature engineering, data quality, bias auditing, and dataset documentation |
| dba model: opus | Database specialist adapting to any engine (PostgreSQL, SQLite, MongoDB, etc.) — schema design, query optimization, migrations, and index tuning |
| devops-engineer model: opus | DevOps engineer for CI/CD pipelines, Docker, PostgreSQL provisioning, monitoring, and deployment |
| documentation-accuracy-reviewer model: inherit | Use this agent when you need to verify that code documentation is accurate, complete, and up-to-date. Specifically use this agent after: implementing new features that require documentation updates, m |
| engineer model: opus | Software engineer specializing in Clean Architecture, SOLID, and root-cause problem solving — adapts to any language and tech stack |
| experiment-runner model: opus | ML experiment design specialist — ablation studies, hyperparameter search, reproducibility, statistical rigor, and result analysis |
| frontend-engineer model: opus | Frontend developer specializing in React/TypeScript with Clean Architecture, component-driven design, and accessibility |
| alexander model: opus | Christopher Alexander reasoning pattern — pattern languages for design knowledge, generative sequences that produce wholeness, decomposition by misfit variables, the fifteen fundamental properties. Do |
| alkhwarizmi model: opus | Al-Khwarizmi reasoning pattern — reduce messy problems to canonical forms via systematic transformation operations (al-jabr and al-muqabala), classify all possible cases exhaustively, then apply known |
| altshuller model: opus | Genrich Altshuller reasoning pattern — systematic invention through contradiction resolution, 40 inventive principles, the Ideal Final Result, evolution patterns of technical systems, resources-in-the |
| archimedes model: opus | Archimedes reasoning pattern \u2014 heuristic-then-proof two-stage discovery, using physics to discover mathematical truths before proving them rigorously, method of exhaustion for bounding by converg |
| arendt model: opus | Hannah Arendt reasoning pattern \u2014 thoughtlessness audit for systemic failures caused by suppressed judgment, labor/work/action classification for evaluating what activities produce durable value, |
| aristotle model: opus | Aristotle reasoning pattern — four-causes interrogation for complete causal explanation, fallacy catalog for argument hygiene, division by differentiae for classification, knowing-that vs knowing-why |
| bateson model: opus | Gregory Bateson reasoning pattern — schismogenesis detection (runaway escalation patterns between interacting parties), double-bind diagnosis (contradictory messages at different logical levels), meta |
| beer model: opus | Stafford Beer reasoning pattern — viable system diagnosis, variety engineering, recursive viability. Domain-general method for diagnosing whether an organization or system can survive in its environme |
| borges model: opus | Jorge Luis Borges reasoning pattern — exhaustive-space audit, map-territory discipline, self-reference detection, forking-paths analysis. Domain-general method for exposing hidden assumptions about co |
| boyd model: opus | John Boyd reasoning pattern \u2014 OODA loop for adversarial decision cycling, orientation as the critical synthesis step, destructive deduction and creative induction for mental model revision, fast |
| braudel model: opus | Fernand Braudel reasoning pattern — three-timescale decomposition of phenomena into structure (longue duree), conjuncture (cycle), and event; structure explains more than events; treat systems as geog |
| bruner model: opus | Jerome Bruner reasoning pattern — the distinction between narrative and paradigmatic (logico-scientific) modes of thought, narrative as a fundamental mode of sense-making distinct from logical argumen |
| carnot model: opus | Sadi Carnot reasoning pattern — deriving theoretical efficiency limits for any process, reversibility audit to find where irreversible losses occur, entropy production localization, comparing actual p |
| champollion model: opus | Jean-François Champollion reasoning pattern — decipherment of unknown representational systems via bilingual bootstrapping, anchor-and-propagate constraint solving, dual-nature recognition. Domain-gen |
| coase model: opus | Ronald Coase reasoning pattern — transaction cost analysis for drawing system/organizational boundaries, build-vs-buy decisions, identifying where the boundary between internal coordination and extern |
| cochrane model: opus | Cochrane/Glass reasoning pattern — systematic evidence synthesis across multiple independent studies, meta-analysis with heterogeneity assessment, publication bias detection, evidence quality grading. |
| curie model: opus | Curie reasoning pattern — instrument-first measurement, name-the-anomaly, and isolate-the-carrier-of-the-residual. Domain-general method for any situation where a signal must be extracted from bulk or |
| darwin model: opus | Charles Darwin reasoning pattern — long-horizon patient observation, systematic collection of variation, a "difficulty book" of observations that contradict your theory, and refusal to publish until t |
| deming model: opus | W. Edwards Deming reasoning pattern \u2014 common-cause vs special-cause variation diagnosis, PDSA cycle with prediction, System of Profound Knowledge, driving out fear for accurate signals. Domain-ge |
| dijkstra model: opus | Edsger W. Dijkstra reasoning pattern — elegance as correctness, proof-and-program developed together, separation of concerns, testing shows the presence not the absence of bugs, and programming as an |
| eco model: opus | Umberto Eco reasoning pattern \u2014 Model Reader/User construction for designing interpretable artifacts, open vs closed design classification, limits of interpretation for detecting overinterpretati |
| einstein model: opus | Albert Einstein reasoning pattern — gedankenexperiment (ride the phenomenon in your mind), operational definitions of abstract concepts, demand covariance (the form of the law must not depend on the o |
| ekman model: opus | Paul Ekman reasoning pattern — anchor subjective experience to observable anatomical units; detect concealed signals via micro-temporal leakage at resolutions the unaided observer misses; establish ba |
| engelbart model: opus | Douglas Engelbart reasoning pattern — augment human capability rather than automate it away; bootstrap (use the tool to improve the tool); the unit of analysis is the co-adapted human-language-artifac |
| erdos model: opus | Paul Erd\u0151s reasoning pattern \u2014 the probabilistic method (prove existence by showing random construction succeeds with positive probability), random graph phase transitions, extremal combinat |
| erlang model: opus | Agner Krarup Erlang reasoning pattern — queuing theory for capacity planning, the nonlinear utilization-latency relationship, Little's Law as a conservation law for queues, blocking probability calcul |
| euler model: opus | Leonhard Euler reasoning pattern \u2014 notation design as infrastructure that makes solutions visible, systematic exhaustive enumeration of structural cases, abstraction by deletion of irrelevant det |
| feinstein model: opus | Feinstein/Sackett reasoning pattern — systematic clinical reasoning through differential diagnosis, Bayesian updating via likelihood ratios, evidence-based practice hierarchy, treatment threshold anal |
| fermi model: opus | Fermi reasoning pattern — order-of-magnitude estimation as a first move, bracket before solve, refuse false precision. Domain-general method for any situation where a precise answer is premature and a |
| feynman model: opus | Richard Feynman reasoning pattern — rederive from scratch to check understanding, explain it to a freshman, detect cargo-cult procedures, and lean over backwards to report what might invalidate your o |
Slash commands (24)
/_index/agent/list/agent/spawn/agent/status/genius/compose/genius/invoke/genius/route/git/clean/git/commit/git/pr/quality/pre-commit/quality/pre-push/research/notebook/research/session/review-pr/session/recall/session/save/skill/run/zetetic/difficulty/zetetic/estimate/zetetic/integrity/zetetic/provenance/zetetic/review/zetetic/verify
Similar rigs
cdeust/zetetic-team-subagents
One epistemic standard on three hosts. Claude Code: 15 problem-shaped skills, 97 sourced reasoning patterns, 120 agents, commit-time gates. Codex and Gemini CLI: portable evidence-synthesis, design and goal-loop skills. Shared gates block u
Orchestrator 6.7k tok · · Q0
VersoXBT/claude-initial-setup
75 skills, 14 agents, 15 commands for Claude Code. Plug-and-play starter kit covering TypeScript, Python, Go, Rust, React, Next.js, FastAPI, Django, Docker, CI/CD, security, testing, and more. Cross-AI support for Cursor, Copilot, and Codex
Orchestrator 5.1k tok · · Q5
piomin/claude-ai-spring-boot
Claude Code template for Spring Boot and cloud native ecosystem
Orchestrator 1.0k tok · · Q0
theishandubey/claude-config
A complete Claude Code setup. An agent roster, an orchestration playbook, safe defaults and curated skills, installed with one command.
Orchestrator 5.0k tok · · Q5