160 specialist AI agents for Claude Code, Cursor & Codex — one npm install puts them all into your IDE as MCP tools and parallel sub-agents. Local-first, BYOK, works offline with Ollama.
This repository also configures its own agents. See what timps-swarm tells them →
Agent Claude Code
Plan and execute an OWASP-style red-team engagement: threat model, adversarial test cases, payloads, runnable exploit harness, and a remediation checklist. Use the timpsredteamagent MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Detect code smells and produce a fully refactored version with a diff summary, complexity scores, and improvement list. Use the timpsrefactoringagent MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Produce a release plan: SemVer bump, release notes, rollout/canary strategy, rollback gate, customer-comms draft, and stakeholder checklist. Use the timpsreleasemanager MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Research Agent — runs BEFORE any coding task. Searches documentation, GitHub issues, Stack Overflow, and the web to gather context so all downstream agents have grounded information. Returns a structured brief: findings, gotchas, recommended approach, sources. Use the timpsresearchagent MCP tool to perform this task.…
Agent Claude Code
Deep-dive research on any topic. Returns a structured brief: key findings, consensus view, controversies, recommended actions, and sources. Depth options: quick, medium (default), deep. Use the timpsresearchscout MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Generate ROS2 nodes, URDF descriptions, motion planning config, and ISO 10218-1 safety analysis. Use the timpsroboticsagent MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Run the full 10-agent SDLC pipeline (product manager → architect → code generator → reviewer → QA → security → performance → devops → docs). Use for software development tasks: writing code, fixing bugs, building APIs, etc. Use the timpsruntask MCP tool to perform this task. Do not answer directly — delegate to this…
Agent Claude Code
Generate a CycloneDX / SPDX Software Bill of Materials from a project tree + manifests, integrate syft/trivy if available, and flag license and supply-chain risks. Use the timpssbomgenerator MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Scan for hardcoded secrets, generate Vault/AWS Secrets Manager migration code and pre-commit hooks. Use the timpssecretsmanagement MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Security scan: open network ports, running process anomalies, app camera/microphone/location permissions (macOS TCC). Returns CVSS-rated findings and mitigation commands. Use the timpssecurityguard MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Convert a security finding (CVE, SAST, secret, misconfig) into a step-by-step patch plan with a working code diff and rollback. Use the timpssecurityremediation MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Score any agent's output on a 1–10 quality scale, identify weaknesses, and automatically re-run the originating agent with an improved prompt until the score meets the threshold. This closes the quality feedback loop across the entire swarm. Use the timpsselfcriticagent MCP tool to perform this task. Do not answer…
Agent Claude Code
Run Lighthouse CI, generate schema.org JSON-LD, fix Core Web Vitals, build sitemaps. Use the timpsseoagent MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Generate production-grade service-mesh config (Istio / Linkerd / Consul) — mTLS, traffic policies, retries, circuit-breakers, observability, multi-cluster, and a migration plan. Use the timpsservicemeshconfigurator MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Decompose epics into sprint tickets with estimates, dependencies, and GitHub Issues payload. Use the timpssprintplanningagent MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Generate a formatted daily standup report from task lists. Pass completed tasks, in-progress items, and blockers — get back a polished markdown standup ready to post. Use the timpssprintreporter MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Analyse a slow SQL query and return: the optimised rewrite, index recommendations with CREATE INDEX SQL, and an estimated speedup. Supports PostgreSQL, MySQL, SQLite, MSSQL. Use the timpssqloptimizer MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Generate an original illustrated children's story for a target age, theme, and moral: outline, full prose, page breaks, illustration prompts, and a printable PDF layout spec. Use the timpsstorybookstorygenerator MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Diagnose why a computer is slow. Scans: top CPU/RAM processes, startup items, thermal throttling, memory pressure. Returns a report with specific processes to kill and startup items to disable. Generates a dry-run cleanup shell script. Use the timpssystemoptimizer MCP tool to perform this task. Do not answer directly…
Agent Claude Code
Scan codebase for TODO/FIXME density, cyclomatic complexity hotspots, and deprecated patterns. Returns a prioritised debt repayment plan. Use the timpstechdebtquantifier MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Refresh a project README, write ADRs, generate a 'Getting Started' guide, an FAQ, and a reference page from the codebase. Use the timpstechwriterassistant MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Quantify technical debt via radon complexity + git churn + TODO density; output a debt roadmap. Use the timpstechnicaldebt MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Review terraform plan output for destructive changes. Requires explicit approval before allowing apply on high-risk plans. Use the timpsterraformplanreviewer MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
Agent Claude Code
Generate realistic, edge-case-covering seed / fixture data for any JSON schema, SQL DDL, or plain-text description. Use the timpstestdataagent MCP tool to perform this task. Do not answer directly — delegate to this sub-agent.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: