Skills
97,222 packaged capabilities and playbooks for agents. Install any of them with one command.
Testing & quality
Analyzes audio recording quality - echo detection, loudness, speech intelligibility, SNR, and spectral analysis. Use when the user wants to check a recording's quality, detect echo or duplication, measure speech clarity, compare original vs processed audio, or diagnose why a recording sounds bad, including tracks from Blackbox or any call recording app.
Audit work, verify claims with concrete evidence, and explain the result in simple grade-5 language. Use when the user asks to review, audit, check, verify, explain a change or fix, summarize test results, validate whether something works, or translate technical findings into plain language for non-technical readers.
Run a rigorous autonomous experiment loop for any optimization target using explicit hypotheses, repeated trials, structured experiment logs, and local HTML reports. Use when asked to "run autoresearch", "optimize X in a loop", "start experiments", or "improve this with benchmark-driven iteration".
Given a research proposal, claim, method, or result, recommend the baselines and benchmarks needed to support it — split into a MINIMAL set (necessary for credibility, skipping any risks rejection) and a SUGGESTED set (complementary, strengthens external validity). Use whenever the user proposes something and asks "what should I compare against", "what baselines do I need", "which benchmarks support this claim", "is this enough to be convincing", "what's the minimal experiment set", "what would a reviewer demand", or shares a method/idea and wants the comparison plan. Default to running this on any proposal so claims are backed by the right evidence. Always check first whether the input is sufficient (precise claim, field, the novel component, metric); if not, request the missing context before recommending rather than guessing. Use experiment-design to design the run itself (or to design a measurement when no benchmark exists), and hypothesis-and-ablation-planner for internal ablations.
Automatic compliance audit against the Hexagone domain business rules (docs/domain/). Analyzes the code of a screen and the APIs it calls, matches them against documented invariants, transitions and validations, and produces a structured report with citations. Report-only mode — no automatic modification of business rules in a healthcare context. Use when the user wants to verify the business compliance of a screen, audit business rules, validate invariants/transitions/validations, or mentions « conformité métier » / « business compliance ».
Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes: Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/spec asked for?). Runs both reviews in parallel sub-agents and reports them side by side. Use when the user wants to review a branch, a PR, work-in-progress changes, or asks to "review since X".
Comprehensive overview of ALL Coval platform resources, their hierarchy, relationships, API endpoints, and ID formats. Use when user asks about Coval resources, data model, how things relate, what endpoints exist, or needs context about the platform structure before making API calls.
This skill should be used when cleaning up codebases that have accumulated dead code, redundant implementations, and orphaned artifacts — especially codebases maintained by coding agents. Triggers on "find dead code", "clean up unused code", "remove redundant code", "prune this codebase", "dead code sweep", "code cleanup", or when a codebase has gone through multiple agent-driven refactors and likely contains overlooked remnants. Systematically identifies cruft, categorizes findings, and removes confirmed dead code with user approval.
Run a design system multi-point inspection — assess a design system across 10 stations (coverage, best practices, accessibility, shared language, testing, orchestration, governance, feedback, machine-readable docs, agent access) and produce a red/yellow/green inspection report with a prioritized work order. Use when the user asks to inspect, assess, audit, or health-check their design system, run the inspection or a specific station, or asks "why is my design system's check engine light on?"
Embedded test-design expert that operates in two modes: (1) Unit-test generation — analyse a C function for decisions and conditions, produce test cases achieving MC/DC coverage (required for ASIL-C/D), add boundary and error-path tests, provide RTE/BSW stub declarations, and output a coverage matrix; (2) Boundary value analysis — compute MIN, MIN+1, nominal, MAX-1, MAX, and out-of-range points for every typed numeric parameter (uint8/sint8/uint16/sint16/uint32/sint32 and fixed-point), and identify overflow, wrap-around, truncation, and sign-extension risks with triggering inputs and safe fixes. Test output is compatible with Unity, CppUTest, and VectorCAST patterns. A third use, characterization testing, pins the existing behaviour of legacy code before a modernization step so equivalence can be proven after the change. Returns decision-ready, self-checked test sets with explicit coverage gaps stated.
Find where an AI/LLM application is failing in production and surface the failure patterns, working from real traces. Use when someone wants to understand what's going wrong with an AI feature, find and categorize failure modes, triage errors, or investigate quality issues (wrong answers, ignored instructions, hallucinations, tool misuse) — "what's failing in my agent", "surface error patterns", "why are the responses bad", "find the common failure modes", "what should I fix next". Covers scoping to one use case, finding failing traces by whichever signal fits the context (code errors, metric outliers, trace-type slices, manual review, existing-eval spikes, clustering), and reading them into a ranked failure taxonomy.
Run Flutter MCP runtime validation from CLI in two steps (launch app, then run validate-runtime), including toolkit-extension gating, screenshot/layout capture, app error collection, optional reload verification, and retry handling for transient first-connect failures.
Multi-engine long-term memory for game AI agents. Deep Godot 4.7+ (99 skills, 27 genre blueprints) + Bevy docs-first parity — audited patterns, persona orchestration (Analyst · Auditor · Builder), production-grade engineering.
Multi-engine long-term memory for game AI agents. Deep Godot 4.7+ (99 skills, 27 genre blueprints) + Bevy docs-first parity — audited patterns, persona orchestration (Analyst · Auditor · Builder), production-grade engineering.
Help a human understand and inspect completed AI-authored repository work through actual results, before-and-after explanations, and supporting evidence. Use when the user asks to understand the work before accepting it, examine assumptions beyond an AI summary, or resume a human review after switching tasks. Also use for explicit requests to judge unresolved product behavior. Ordinary change summaries, automated defect hunting, and non-repository content do not need this skill.
Implements a saved plan through to a tested, independently reviewed pull request, handing it to an agent in its own worktree when possible. Use this skill when the user asks to implement, execute, or continue a plan. Do not use for creating a plan (use create-plan) or for changes without a plan.
Use when writing or reviewing tests for external API adapters, storage adapters, database integration boundaries, or service contracts. Covers fixture design, schema validation, negative contracts, fakes, recorded responses, and CI-safe verification. Do not use for simple pure unit tests with no boundary contract.
Guide the ZenML Pro hosted Kitaru onboarding tour from preloaded traces through human review, one reusable evaluator, and one bounded replay. Use only inside the controlled hosted onboarding runner. Resume exact durable workspace state, handle existing agents without name collisions, and keep the tour concise.
After navigating and interacting in Cursor's built-in browser, use browser_network_requests to audit every fetch/XHR for failures, slowness, duplicate calls, and suspicious payloads. Use for API-heavy pages and after backend or client networking changes.
Mandatory pre-code gate and definition-of-done for ANY Odoo change (add field, override method, inherit view/xpath, OWL/JS patch, wizard, cron, controller, report, security, migration, bug fix, refactor, "optimize", "clean code", "stick to base"), Odoo 16-19. Also when the user pastes an Odoo traceback or install/upgrade error (ParseError "cannot be located in parent view", "Field ... does not exist", missing xmlid) or asks where an Odoo model, field, method, view or xmlid is defined. Run BEFORE brainstorming or planning: Odoo design questions are answered by tracing the real source into a Context Brief (file:line), not by open-ended Q&A or memory. Resolves the Odoo runtime and version, loads the matching odoo-<major> pack, and closes with install, test, upgrade, i18n, security and hygiene checks before commit.
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.