5,399 tagged ai-agents, measured the same way as everything else here.
Browse within: agentic-ai 330LLM 299ai-coding 185prompt-engineering 143agent-orchestration 133agentic 128ai-software-team 125claude-code-plugin 113cli 109agents-md 100Science 99ai-skills 99agent-framework 86agentic-coding 74
pedrobraiti/vizier-trading-skill
Agent Claude Code
Research-only market envoy for Vizier's breadth-discovery (manager) mode. Sweeps ONE coverage area of the market with Scout data tools and returns a structured candidate shortlist. It has NO execution tools — it cannot place an order — so it is safe to fan out across the market in parallel. Spawned by the Vizier…
Agent Claude Code
Deep implementation work delegated by the lead — writing kit code, writing tests (five-exit-doors discipline), debugging failures, and the implementer self-review pass on a diff. Use for any coding work bigger than a trivial edit. Runs on Opus.
Agent Claude Code
Software engineer specialist (IRC handle @dev) dispatched to write and test application code, database schema migrations, and infrastructure configuration files.
Agent Claude Code
defi-engineer.md, solana-architect.md, and solana-qa-engineer.md are adapted from solanabr/solana-ai-kit by Superteam Brazil, used under the MIT License. They give this starter a trading brain (Jupiter/Pyth/oracle-guard patterns), a settlement-program architecture voice (escrow/PDA design), and a Solana test…
Agent Claude Code
Use for any Brain OS task delegated to a subagent — reading entity state, checking active decisions, running focus/patterns/retro analysis, or proposing changes that need to honor active decisions. Always uses Brain OS MCP tools (mcpbrain-os) first; falls back to file reads only when the MCP server is unreachable.…
Agent Claude Code
Standalone accessibility design reviewer evaluating ARIA patterns, focus management, state communication, and multi-perspective access (screen reader, keyboard, low vision, cognitive). 8-phase investigation protocol with strict evidence requirements and calibrated severity ratings.
Agent Claude Code
Write external-facing DevRel content — Twitter/X threads, LinkedIn posts, YouTube video scripts, podcast outlines, conference talk abstracts, hackathon submission narratives, customer-facing announcements. Use this subagent when the user asks to "draft a thread", "write a script for the video", "outline the talk"…
Agent Claude Code
Demo worker-shaped orchestrator. Dispatches to implementer / reviewer / qa specialists per task class. Modeled on the real worker.md so that slice-1 q8rl regex (literal subagenttype) and slice-2 hw6j frontmatter (placeholder declaration) both have non-Claude-only fixture coverage.
lhhiep2204/apple-agent-builder-kit
Agent Claude Code
Generate Apple-platform Copilot agents, skills, instructions, prompts, templates, and hook guidance from user requirements and codebase analysis for implementation, review, testing, architecture, and agile delivery workflows.
Agent Claude Code
A coding role that turns detailed plans into code, tests the result, and fixes problems found during review.
Dankosik/go-service-template-rest
Agent Claude Code
Mutable worker for one bounded implementation, investigation, or verification result.
Agent Claude Code
Leads an end-to-end user task, deciding the next useful action and integrating verified results. Use when one agent owns the final outcome.
Agent Claude Code
Analyzes type design for encapsulation quality, invariant expression, usefulness, and enforcement, providing quantitative ratings and actionable improvement suggestions.
Agent Claude Code
Visual QA specialist using Playwright MCP and Figma MCP for pixel-perfect frontend implementation and debugging.
Agent Claude Code
Guardian of redbar's design and the 7 rules in AGENTS.md. Invoke BEFORE changes that risk an invariant (a new effect boundary, a new public field, the way a language is added, an output format), to review architectural drift, to write or review specs in docs/superpowers/specs/, and when a design decision must be made…
Agent Claude Code
Equity-compensation tax planning. Use proactively for incentive stock option (ISO) or alternative minimum tax (AMT) exercise timing, non-qualified stock option (NSO) tax, restricted stock unit (RSU) vesting, Section 1202 Qualified Small Business Stock (QSBS) qualification, single-stock concentration, protective-put or…
Agent Claude Code
General-purpose software engineer. Picks up implementation tickets, writes and refactors code, runs tests, and closes with evidence. Good default for any ticket without a more specialized agent. Triggers: any TASK ticket tagged agent=engineer, or when no other agent is a better match.
Agent Claude Code
Software architecture specialist for system design, scalability, and technical decision-making. Use PROACTIVELY when planning new features, refactoring large systems, or making architectural decisions.
curtischoutw/claude-institution
Agent Claude Code
A review agent that examines code or designs for security risks and ways they may fail. It checks inputs, permissions, secrets, concurrent operations, partial failures, and injection attacks, where untrusted data changes a command or query.
Agent Claude Code
GLM advocate for strategic debates - presents alternative perspective via GLM API.
bhuiyanmobasshir94/claude-standing-orders
Agent Claude Code
Read-only code reviewer. Use after implementation work and before integration, especially on multi-file changes, refactors, and anything touching auth, credentials, migrations, money, PII, or public API shape. Reads the diff and measures it against the project's stated standards, citing file:line for every finding. It…
Agent Claude Code
Fresh-context Quality Loop reviewer. Use after implementation and before package/done.
Agent Claude Code
Use when the user asks to audit memory-bank hygiene — checking for stale status: active files in tasks/work/, broken @-references in CLAUDE.md / skills, memory-bank files missing from index.md, frontmatter inconsistencies, or untracked files. Reports findings only; never writes.
IncommensurableHubris/fullstack-director
Agent Claude Code
Fullstack Director's blind, read-only OWASP panel reader (skill 07). Each spawn owns exactly ONE area-slice (given in the prompt — classic R1–R4/R5 or the agent-system flipped partition), analyzes only its slice with a neutral evidence-required stance, may run that area's deterministic scanners, and returns findings…
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: