Agent Claude Code
Python/FastAPI backend expert. Use when implementing or changing API endpoints, services, DB access, or Celery workers. Enforces thin routers, service layer, typing, and tests.
not rated 0 1mo ago A 42 tokens
debugger
02Agent Claude Code
Debugging specialist for errors and test failures. Use when encountering issues.
not rated 0 1mo ago A 17 tokens
devops
03Agent Claude Code
DevOps/infra specialist. Use when working on deployment, environments, monitoring (e.g. Prometheus/Alertmanager), infra scripts, or secrets. Covers deploy, env config, and monitoring maintenance.
not rated 0 1mo ago A 44 tokens
Agent Claude Code
Documentation and runbook keeper. Use when updating docs, README, or script descriptions; prefer updating existing canonical docs over creating new analysis/report files.
not rated 0 1mo ago A 33 tokens
hygiene
05Agent Claude Code
Post-work hygiene: merge analyses into canonical docs, remove temp files, propose cleanup. Use at task end or when user asks to tidy. Align with /cleanup and migration-and-doc-consolidation.
not rated 0 1mo ago A 0 tokens
refactorer
06Agent Claude Code
Refactoring and SOLID specialist. Use when splitting responsibilities, introducing abstractions (DB/cache), removing duplication, or applying SOLID to existing code. Prefer small, test-backed steps.
not rated 0 1mo ago A 41 tokens
rules-keeper
07Agent Claude Code
Maintains project rules: audit .cursor/rules/, prefix convention (cursor-rules-principles), sync doc/rules.md. Use when updating or auditing project rules.
not rated 0 1mo ago A 0 tokens
Agent Claude Code
Security specialist. Use proactively when implementing auth, payments, or handling sensitive data.
not rated 0 1mo ago A 21 tokens
test-runner
09Agent Claude Code
Test automation expert. Use proactively to run tests and fix failures.
not rated 0 1mo ago A 17 tokens
verifier
10Agent Claude Code
Validates completed work. Use after tasks are marked done to confirm implementations are functional.
not rated 0 1mo ago A 20 tokens
cleanup
11Command Claude Code
You are in CLEANUP mode. Goal: restore repo order after agent work. Default is SAFE (dry-run). Do NOT delete/move/rename until user confirms: APPLY CLEANUP.
not rated 0 1mo ago A 0 tokens
next
12Command Claude Code
You are in NEXT mode. Goal: (1) check project status, (2) pick the right next task, (3) switch to the appropriate working mode, (4) begin execution immediately when the next task is clear.
not rated 0 1mo ago A 0 tokens
Command Claude Code
You are in PONYTAIL REVIEW mode. Goal: find code that didn't need to be written — unnecessary abstractions, unused dependencies, dead code, premature generalization — without touching validation, error handling, security, or accessibility.
not rated 0 1mo ago A 0 tokens
Command Claude Code
Tools (user-memory MCP): Use the tools that actually exist. There is no memoryfind or databaseswitch in this MCP.
not rated 0 1mo ago A 0 tokens
retro
15Command Claude Code
Context: Analyze THIS agent chat thread (full conversation visible in Cursor).
not rated 0 1mo ago A 0 tokens
save_memory
16Command Claude Code
Store the following as persistent memory NOW (no extra confirmation).
not rated 0 1mo ago A 0 tokens
status
17Command Claude Code
You are in STATUS mode. Goal: report where the project stands and what should happen next — without editing files, switching modes, or starting work.
not rated 0 1mo ago A 0 tokens
browserclaw
18Skill Claude CodeCodex
The user's dedicated browser for agents — a real browser signed into their accounts, with live logins and a persistent profile. Use it for any task that touches a website or browser (open, read, act, fill, sign in, download, verify). The user installed it precisely so agents default here unprompted — over in-app…
not rated 0 1mo ago A 85 tokens
Skill Claude Code
Ensure commands and paths are safe across Windows + WSL + Docker (PowerShell vs bash, path translation, permissions). Use when instructions touch shell, filesystem, Docker, or OS-specific behavior.
not rated 0 1mo ago A 44 tokens
Skill Claude CodeCodex
Generate Cursor deeplinks (prompt/command/rule and MCP install links) using the repo generator and the project output contract. Use when the user asks for a Cursor deeplink, cursor:// links, cursor.com/link links, or wants to create prompts/commands/rules via deeplinks.
not rated 0 1mo ago A 66 tokens
Skill Claude CodeCodex
Implement or refactor data import/parsing so it streams files sequentially (memory-safe), validates/coerces types explicitly, skips irreparable records while logging them to an error CSV (full original columns + timestamp/file/line/error), emits rowsok/rowsskipped/parseerrors metrics, and guarantees idempotent DB…
not rated 0 1mo ago A 72 tokens
Skill Claude Code
Conduct deep, source-grounded research with freshness checks and citations, then synthesize an actionable answer. Use for time-sensitive, niche, or high-stakes topics; when the user asks to “zweryfikuj”, “zrób research”, or when uncertainty is high.
not rated 0 1mo ago A 61 tokens
Skill Claude Code
Apply enhanced verification (CoVe-like review, security/arch/perf checks, targeted web research) for high-risk tasks or uncertainty. Use for security, infra, data loss risk, major refactors, or when facts may be outdated.
not rated 0 1mo ago A 52 tokens
Skill Claude CodeCodex
Apply frontend/UI design guardrails and the Impeccable anti-pattern checklist (pbakaus/impeccable) when building or reviewing visual UI — landing pages, dashboards, marketing sites, component styling. Use for design critique, polish passes, or before shipping new UI; pairs with vanilla-web/next-stack (implementation)…
not rated 0 1mo ago A 77 tokens
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: