noctua84/nescio-ai
An agent crew for Claude Code that says 'I don't know' — principled refusal + a built-in devil's advocate, a growing memory, and a learning loop.
field: AI / MLSecurity workflow: Planning & specsSecurity reviewFrontend & designPR / code reviewTesting & TDD
stack: Python
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 noctua84/nescio-ai/agents ./rig-nescio-ai/agents $ npx degit noctua84/nescio-ai/commands ./rig-nescio-ai/commands $ npx degit noctua84/nescio-ai/skills ./rig-nescio-ai/skills $ npx degit noctua84/nescio-ai/hooks ./rig-nescio-ai/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 noctua84/nescio-ai/skills/adopt-config .claude/skills/adopt-config $ npx degit noctua84/nescio-ai/skills/agent-evaluation .claude/skills/agent-evaluation $ npx degit noctua84/nescio-ai/skills/api-security-assessment .claude/skills/api-security-assessment $ npx degit noctua84/nescio-ai/skills/code-navigation .claude/skills/code-navigation $ npx degit noctua84/nescio-ai/skills/compliance-gap-analysis .claude/skills/compliance-gap-analysis $ npx degit noctua84/nescio-ai/skills/create-adr .claude/skills/create-adr $ npx degit noctua84/nescio-ai/skills/dependency-pr-ci-fix .claude/skills/dependency-pr-ci-fix $ npx degit noctua84/nescio-ai/skills/detection-rule .claude/skills/detection-rule $ npx degit noctua84/nescio-ai/skills/fix-security-vulnerabilities .claude/skills/fix-security-vulnerabilities $ npx degit noctua84/nescio-ai/skills/gh-milestones-projects .claude/skills/gh-milestones-projects $ npx degit noctua84/nescio-ai/skills/handle-pr-comments .claude/skills/handle-pr-comments $ npx degit noctua84/nescio-ai/skills/hipaa-assessment .claude/skills/hipaa-assessment $ npx degit noctua84/nescio-ai/skills/incident-response-plan .claude/skills/incident-response-plan $ npx degit noctua84/nescio-ai/skills/ir-playbook .claude/skills/ir-playbook $ npx degit noctua84/nescio-ai/skills/iso27001-isms .claude/skills/iso27001-isms $ npx degit noctua84/nescio-ai/skills/layered-api-design .claude/skills/layered-api-design $ npx degit noctua84/nescio-ai/skills/llm-application-architecture .claude/skills/llm-application-architecture $ npx degit noctua84/nescio-ai/skills/modular-design .claude/skills/modular-design $ npx degit noctua84/nescio-ai/skills/pci-dss-assessment .claude/skills/pci-dss-assessment $ npx degit noctua84/nescio-ai/skills/prompt-engineering-guide .claude/skills/prompt-engineering-guide $ npx degit noctua84/nescio-ai/skills/prompt-evaluation-harness .claude/skills/prompt-evaluation-harness $ npx degit noctua84/nescio-ai/skills/prompt-testing-plan .claude/skills/prompt-testing-plan $ npx degit noctua84/nescio-ai/skills/publish-adrs .claude/skills/publish-adrs $ npx degit noctua84/nescio-ai/skills/rag-system-design .claude/skills/rag-system-design $ npx degit noctua84/nescio-ai/skills/repo-hygiene .claude/skills/repo-hygiene $ npx degit noctua84/nescio-ai/skills/risk-assessment .claude/skills/risk-assessment $ npx degit noctua84/nescio-ai/skills/risk-register .claude/skills/risk-register $ npx degit noctua84/nescio-ai/skills/sbom .claude/skills/sbom $ npx degit noctua84/nescio-ai/skills/sdlc-security-gates .claude/skills/sdlc-security-gates $ npx degit noctua84/nescio-ai/skills/secure-code-review .claude/skills/secure-code-review $ npx degit noctua84/nescio-ai/skills/secure-design-spec .claude/skills/secure-design-spec $ npx degit noctua84/nescio-ai/skills/security-architecture-review .claude/skills/security-architecture-review $ npx degit noctua84/nescio-ai/skills/soc2-report .claude/skills/soc2-report $ npx degit noctua84/nescio-ai/skills/threat-model .claude/skills/threat-model $ npx degit noctua84/nescio-ai/skills/vuln-assessment .claude/skills/vuln-assessment $ npx degit noctua84/nescio-ai/skills/zero-trust-design .claude/skills/zero-trust-design
$ curl -fsSL --create-dirs -o .claude/agents/advisor.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/advisor.md $ curl -fsSL --create-dirs -o .claude/agents/builder-simple.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/builder-simple.md $ curl -fsSL --create-dirs -o .claude/agents/builder-standard.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/builder-standard.md $ curl -fsSL --create-dirs -o .claude/agents/builder.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/builder.md $ curl -fsSL --create-dirs -o .claude/agents/critic.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/critic.md $ curl -fsSL --create-dirs -o .claude/agents/doc-researcher.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/doc-researcher.md $ curl -fsSL --create-dirs -o .claude/agents/doc-writer.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/doc-writer.md $ curl -fsSL --create-dirs -o .claude/agents/explore.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/explore.md $ curl -fsSL --create-dirs -o .claude/agents/librarian.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/librarian.md $ curl -fsSL --create-dirs -o .claude/agents/orchestrator.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/orchestrator.md $ curl -fsSL --create-dirs -o .claude/agents/planner.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/planner.md $ curl -fsSL --create-dirs -o .claude/agents/qa-guard.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/qa-guard.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/scout.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/scout.md $ curl -fsSL --create-dirs -o .claude/agents/test-writer.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/test-writer.md $ curl -fsSL --create-dirs -o .claude/agents/validator.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/validator.md $ curl -fsSL --create-dirs -o .claude/agents/vision.md https://raw.githubusercontent.com/noctua84/nescio-ai/main/agents/vision.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 *)"
]
}
} Skills (36)
adopt-configagent-evaluationapi-security-assessmentcode-navigationcompliance-gap-analysiscreate-adrdependency-pr-ci-fixdetection-rulefix-security-vulnerabilitiesgh-milestones-projectshandle-pr-commentshipaa-assessmentincident-response-planir-playbookiso27001-ismslayered-api-designllm-application-architecturemodular-designpci-dss-assessmentprompt-engineering-guideprompt-evaluation-harnessprompt-testing-planpublish-adrsrag-system-designrepo-hygienerisk-assessmentrisk-registersbomsdlc-security-gatessecure-code-reviewsecure-design-specsecurity-architecture-reviewsoc2-reportthreat-modelvuln-assessmentzero-trust-design
Subagents (17)
| advisor model: claude-opus-5 | Read-only architecture advisor. Deep reasoning for debugging, design decisions, and multi-system tradeoffs. |
| builder-simple model: claude-haiku-4-5 | Implementation specialist — simple tier. Mechanical, unambiguous tasks where the pattern to copy is already named and there is no decision to make. Same contract as builder; runs on Haiku for cost eff |
| builder-standard model: claude-sonnet-5 | Implementation specialist — standard tier. Tasks needing some judgment: one or two local decisions, or an existing pattern that needs adapting rather than copying. Same contract as builder; runs on So |
| builder model: claude-opus-5 | Implementation specialist. Executes one scoped task from a plan — writes the code, proves it works, reports honestly. Complex tier — design judgment: architectural decisions, cross-system impact, or g |
| critic model: claude-opus-5 | Devil's-advocate reviewer. Challenges a plan's approach and assumptions in a single bounded pass — blind spots, shaky premises, overlooked alternatives, and PII/legal exposure — then returns ranked ch |
| doc-researcher model: claude-sonnet-5 | Documentation landscape specialist. Reads existing project docs and returns a structured map of coverage, gaps, and update targets — given a description of what just changed. Produces findings for doc |
| doc-writer model: claude-sonnet-5 | Documentation author. Consumes doc-researcher findings and a description of what changed, then writes or updates documentation files using the project's existing vocabulary and structure. Hard file bo |
| explore model: claude-haiku-4-5 | Codebase search specialist. Finds files, code patterns, and relationships. Fire multiple in parallel for broad searches. |
| librarian model: claude-sonnet-5 | External documentation and OSS code search specialist. Finds official docs, implementation examples, and best practices using web search and GitHub. |
| orchestrator model: claude-opus-5 | Full-lifecycle development coordinator. Manages discovery, analysis, planning, execution, review, and delivery by dispatching specialized agents. Guides the user through approval gates between phases. |
| planner model: claude-opus-5 | Strategic planning consultant. Interviews users, gathers requirements, and creates detailed work plans. Read-only except for .sisyphus/ markdown files. |
| qa-guard model: claude-sonnet-5 | CI gate specialist. Discovers the project's CI checks from config files, runs them, fixes mechanical failures (formatting, linting, type annotations, test setup), and iterates until all checks pass or |
| reviewer model: claude-opus-5 | QA engineer. Audits implemented code or features for bugs, regressions, security flaws, and quality issues, then files a dated, severity-ranked report. Works on code whether pre-delivery (in a worktre |
| scout model: claude-opus-5 | Pre-planning consultant. Analyzes requests to identify hidden intentions, ambiguities, and AI failure points before planning begins. |
| test-writer model: claude-opus-5 | Test authorship specialist. Writes and extends tests for implemented code — verifies the intended interface, not the current output. Hard file boundary: may not touch implementation files under any ci |
| validator model: claude-opus-5 | Work plan reviewer. Verifies plans are executable with valid references. Blocker-finder, not perfectionist. |
| vision model: claude-sonnet-5 | Interprets media files (PDFs, images, diagrams, screenshots) that require visual analysis beyond what the Read tool extracts as plain text. Returns only the extracted information, not the raw file. |
Slash commands (2)
/assess-repo/harvest-memory
Plugins (3)
superpowers@claude-plugins-officialtypescript-lsp@claude-plugins-officialpyright-lsp@claude-plugins-official
Permissions
deny (0)
—
ask (0)
—
allow (15)
Bash(find:*)
Bash(grep:*)
Bash(cat:*)
Bash(ls:*)
Bash(echo:*)
Bash(test:*)
Bash(tree:*)
Bash(git status:*)
Bash(git diff:*)
Bash(git add:*)
Bash(git commit:*)
Bash(git checkout:*)
Bash(git mv:*)
WebSearch
WebFetch(domain:www.npmjs.com)
Similar rigs
fjerbi/claude-code-arsenal
A tested Claude Code operating system for efficient, reliable, and focused software engineering. Includes global instructions, workflows, rules, and tools designed to minimize wasted tokens, prevent hallucinations and unnecessary loops, enf
YOLO Cowboy 4.0k tok · · Q51
Warshoow/agent-setup
Documented snapshot of my Claude Code setup on WSL: global config, custom agents and skills, hooks, plugins. An archive, not an installer.
Orchestrator 2.3k tok · · Q0
AkshatNaruka/aidentity
A verified index of 526 AI agent skills — licenses, sources, one-command install for Claude Code, Codex, Cursor, OpenCode, Copilot and Gemini CLI.
Orchestrator 2.6k tok · · Q5
Gildaciolopes/agentic-dev-stack
Smart vibe coding with Claude Code: multi-agent orchestration (Opus plans; Sonnet, Haiku, Gemini, Codex and free models execute), Spec Kit SDD, caveman + RTK + graphify token savings, sandboxed secrets, self-healing free models and a Hermes
Minimalist 1.2k tok · · Q10