agentic-os

Use when adversarially reviewing a document that makes formal or technical claims — math derivations, physics papers, statistical analyses, benchmark reports, whitepapers. Inventories every equation and quantitative claim, verifies each AS NAMED in the text (never a paraphrase or a neighboring statement), and classifies VERIFIED / REFUTED / UNVERIFIABLE / VACUOUS. Triggers on "check this paper", "verify these claims", "is this derivation right", "review this proof", "audit this benchmark", "does the math hold up". For source-code review see code-review-and-quality; for skill/agent library audits see skill-library-review; for content quality scoring see content-ops.

Run Karpathy-style autoresearch optimization on any content. Generates 50+ variants, scores with a 5-expert simulated panel, evolves winners through multiple rounds, outputs optimized version + full experiment log. Use when optimizing landing pages, email sequences, ad copy, headlines, form pages, CTA text, or any conversion-focused content. Triggers on "optimize this page", "run autoresearch", "score these variants", "A/B test this copy".

Tests in real browsers. Use when building or debugging anything that runs in a browser. Use when you need to inspect the DOM, capture console errors, analyze network requests, profile performance, or verify visual output with real runtime data via Chrome DevTools MCP.

Conducts multi-axis code review across correctness, readability, architecture, security, and performance. Use before merging any change. Use when reviewing code written by yourself, another agent, or a human. Triggers on "review my PR", "review this diff", "code review", "review this changeset", "is this ready to merge", "pre-merge review".

Estimate the full development cost of an existing codebase from lines of code, architectural complexity, and team-composition overhead. Use for 'how much would this cost to build', 'what did this codebase cost', development-cost or build-cost estimates, calendar-time estimates, and Claude/AI ROI on a delivered codebase. Estimates by measured LOC and complexity, not by ticket volume.

Auto-assembles a domain-specific expert panel (7–10 experts), scores any content or strategy artifact against a typed rubric, and iterates until the aggregate hits 90+ (max 3 rounds). Use as a quality gate on copy, email sequences, landing-page drafts, strategy docs, charts, titles, or recruiting evaluations — or when another skill needs a final review gate on its output. Triggers on "expert panel this", "score this", "rate these variants", "quality check this", "panel review", "expert score", "evaluate this copy/strategy/page". For variant generation and multi-round conversion optimization see autoresearch; for live-URL CRO auditing see conversion-ops; for the scripted content-production pipeline see content-pipeline.

Non-interactive content-production toolkit: mine quotable moments from podcast RSS feeds and meeting notes, discover clip-worthy moments in video transcripts, repurpose long-form source into platform-native drafts (X, LinkedIn, YouTube Shorts, newsletter), and batch-score/gate those drafts before publish. Use when asked to "mine quotes from this podcast", "find clips in this video", "repurpose this into a thread / LinkedIn post / Short", "turn this transcript into posts", "extract viral moments", or "gate this batch of drafts". Runs Python scripts end to end. For interactive expert-panel scoring of a single artifact see content-ops.

AI-powered conversion rate optimization: landing page audits, CRO scoring, survey segmentation, and lead magnet generation. Use when asked to audit a landing page, run a CRO analysis, segment survey responses, or generate lead magnets. For content quality scoring see content-ops.

Catalog APIs, persistence models, and message/event payloads into DATA_MODEL.md at the project root. Use after implementation when a diff touches request/response types, schemas, ORM models, queue payloads, or webhook shapes. Triggers on "document data model", "update DATA_MODEL", "catalog API shapes".

Adversarially verifies DATA_MODEL.md property rows against cited Source files. Use after data-model-documenter when DATA_MODEL.md changed — inventories each property in added/changed catalog sections and classifies VERIFIED / REFUTED / UNVERIFIABLE. Triggers on "verify DATA_MODEL", "check catalog against source", "data model verification".

Use when authoring or reviewing CI/CD pipelines — GitHub Actions workflows, reusable workflows, composite actions, OIDC federation to AWS/GCP, caching, artifacts, and pipeline security hardening. Triggers on edits to .github/workflows/**, action.yml, composite action definitions, or mentions of "CI", "CD", "pipeline", "GitHub Actions", "workflow", "OIDC", "runner", "deploy script", "release", or "build pipeline".

Use when recording, tallying, or triaging stochastic review findings — the append-only JSONL ledger where Tier 2 (advisory, unevidenced) reviewer output goes instead of blocking language. Triggers on "log this finding", "ledger add", "tally the findings", "triage findings", "is this finding recurring", "retire this noise", or whenever a reviewer produces an unevidenced concern that should be measured for recurrence rather than chased. For the review itself see code-review-and-quality or skill-library-review; for verifying formal claims with deterministic evidence see adversarial-claims-reviewer.

Use when tuning the numbers in a game — economy curves, progression rates, difficulty pacing, drop tables, win/loss odds, time-to-X targets, and currency velocities. Triggers on "balance pass", "tune the economy", "progression curve", "difficulty curve", "drop rate", "XP curve", "TTK", "time-to-content", "balance the economy", "spreadsheet model", "economy sim", or a system spec from game-systems-designer with `TBD` placeholders. Produces an economy spreadsheet, balance pass plan, sim outputs, and per-system balance tables. Stops at the numbers — does not change system rules or pricing. For systems design see game-systems-designer; for store catalog see iap-manager.

Use when designing the systems of a game from a locked concept — core loops, meta loops, player verbs, progression, content systems, level structure, and narrative integration. Triggers on "game design doc", "GDD", "system spec", "core loop design", "progression design", "level design", "MDA", or "design the systems". Produces a design doc plus per-system specs that game-balancer numbers, iap-manager stocks, and godot-engineer builds. Stops at design — no number tuning, pricing, or engine code. For balance numbers see game-balancer; for the engine implementation see godot-engineer.

Use when building games or interactive software in Godot 4 with C# — designing scenes and nodes, writing gameplay code, handling input, physics, animation, UI, rendering, save systems, performance work, WebSocket-based multiplayer, or exporting to platforms. Triggers on "Godot", "GDScript", "C# Godot", "scene tree", Godot node types (Node2D, Node3D, Control, CharacterBody), "_Process", "_PhysicsProcess", "signal", "autoload", "RPC", "WebSocketMultiplayerPeer", "shader", "export preset", or any Godot 4 project files. For game design see game-systems-designer; for balance see game-balancer; for game-specific intake see game-design-shaper.

Autonomous growth experimentation framework: creates experiments with hypotheses, logs data points, runs statistical analysis (bootstrap CI + Mann-Whitney U), auto-promotes winners to a living playbook, and suggests next experiments. Use when asked to run growth experiments, analyze A/B tests, build experiment scorecards, or generate pacing alerts. For SEO-specific experiments see seo-ops; for outbound experiments see outbound-engine.

Use when designing and operating the in-app purchase catalog of a game — SKU design, price-tier ladder, bundles, starter packs, battle pass tiering, A/B price tests, store config (App Store / Google Play / Steam DLC / web / web3), and per-region price localization. Triggers on "IAP catalog", "store SKUs", "pricing tiers", "starter pack", "bundle design", "battle pass tiering", "price test", "price localization", "App Store Connect", "Google Play Console", "Steam DLC", or when handed a monetization strategy with the catalog still open. Produces a catalog spec, per-region price tables, bundle compositions, and a price-test plan. For in-game economy curves see game-balancer.

Use when forensically auditing a Claude Code skill/agent/command/workflow library against RULESET.md by a fixed mechanical protocol — probing every file against the mechanically-checkable rules and reporting CONFORMS / VIOLATES / UNVERIFIABLE / N-A counts with quoted evidence. It casts NO judgment and emits NO pass/fail verdict; counts are the headline. Triggers on "investigate the library", "audit against RULESET", "find every violation", "forensic library audit", "evidence-only check". Not for routing/quality/single-responsibility judgment — use skill-library-review. Not the full sharded sweep — use the audit-library command.

Use to structure a vague marketing request into a well-scoped brief before any real work begins. Triggers on "shape this campaign", "plan this launch", "scope this content", "marketing plan", "growth plan", "content strategy", "outbound plan", or when invoked as the /mshape slash command. Produces a filled brief (campaign, content, optimization, research, or pipeline) that downstream marketing skills can act on. Do not use for work already well-defined — go straight to execution. For engineering task shaping see prompt-shaper; for game-design intake see game-design-shaper.

End-of-session pass that persists durable facts from the CURRENT session into .claude/memory/. Runs IN-SESSION in the main agent (never via Task — a subagent starts cold with no transcript), prompted by the Stop hook's nudge. Triggers on "persist durable facts from this session", "run memory-extraction", "flush session memory", or "extract memory before close". Reads the in-context transcript plus existing memory, applies the durable-fact predicate below, and writes one file per fact plus one MEMORY.md index line — append-or-update, never clobbering a consumer's existing memory.

Design, analyze, and optimize cold outbound email campaigns on Instantly. Handles end-to-end ICP definition, expert panel scoring (recursive to 90+), sequence copywriting, infrastructure audit, capacity planning, and an implementation/strategy doc for human review. Use when asked to build cold outbound sequences, optimize cold email, audit an outbound motion, write sales sequences, or design cold email campaigns. Supports "start from scratch" and "optimize existing" modes. Execution skill — it produces copy, math, and a doc; it does not auto-send. Requires Instantly for audit/send features.

Use when building games or interactive software in Phaser 3 with TypeScript — scenes, gameplay code, input, Arcade physics, animations, audio, tweens, asset preloading, save systems, performance work, or scaffolding with Vite. Triggers on "Phaser", "Phaser 3", "Phaser.Scene", "Phaser.Game", "Arcade physics", "Matter physics", "tilemap", "Tiled", "GameObject", "Container", "Group", "Vite + Phaser", or any `.ts`/`.js` file in a Phaser project. For game design and balance see game-systems-designer, game-balancer; for game intake see game-design-shaper. For Godot/C# see godot-engineer.

Breaks work into ordered, parallel-dispatchable tasks with an execution DAG. Output format is consumable by CI matrices and parallel agent runners — each task has a stable ID, declared file writes, conflict edges, and branch suffix. Use when you have a spec, brief, or shaper output and need to decompose it into implementable units. Use when a task feels too large, when scope spans multi-repo or multi-week work, or when parallel execution across multiple agents is on the table.

Structures a vague engineering request into a well-scoped task brief before any implementation begins. Use when the user has an engineering goal but the ask is missing which repos are in scope, what "done" means, constraints, or open questions. Triggers on "shape this", "scope this out", "frame this work", "write a brief for", "I want to build" (with unclear scope), or the /shape slash command. Produces a filled task template (multi-repo feature, single-repo change, investigation, or bugfix). Not for already-scoped work — go straight to execution. If the domain is unclear, ask one qualifying question before routing — for marketing intake see marketing-shaper, game-design intake see game-design-shaper.

Coordinates release preparation for your monorepo — maintaining CHANGELOG and the release assessment document, resolving merge conflicts on release branches, and communicating status, risk, and asks to the broader team. Use when the user mentions release manager, release train, cut a release, release branch, CHANGELOG, release assessment, monorepo release, merge conflicts during release, versioning, or coordinating a version bump with engineering and stakeholders.

AI-powered revenue intelligence: sales call insight extraction, content-to-revenue attribution, and multi-source client reporting. Use when asked to analyze sales calls, build revenue attribution models, or generate client reports.

Use when writing, reviewing, or architecting Rust code — systems programming, async services, CLI tooling, web backends, or any work in `.rs` files. Triggers on editing `.rs` or `Cargo.toml` files, or mentions of "Rust", "Tokio", "Axum", "cargo", "borrow checker", "lifetime", "trait object", "async Rust", "crate", "rustc", or explicit requests to "build this in Rust" / "rewrite X in Rust". For adversarial security review of Rust code see security-reviewer.

Cross-stack security review — auditing vulnerabilities across infrastructure, smart contracts, CI/CD pipelines, and AI agent systems, plus auth/sessions/crypto and validating user input at API and infrastructure boundaries. Triggers on mentions of "vulnerability", "pentest", "OWASP", "access control", "injection", "CSRF", "JWT", "smart contract audit", "supply chain", "OIDC", or any review of security-sensitive code paths spanning more than one layer. For PII sanitization see [security](../security/SKILL.md).

Scan and redact PII and sensitive data (emails, phone numbers, SSNs, API keys, IP addresses, credentials, amounts, company/person names) from repository files. Includes a pre-commit hook to block commits containing PII. Use when asked to audit code for sensitive data, sanitize files before publishing, or install PII detection hooks. For application security hardening see security-engineering.

AI-powered SEO operations: keyword intelligence, competitor gap analysis, Google Search Console optimization, and trend detection. Use when asked to research keywords, analyze competitor content gaps, audit GSC performance, or detect trending topics. For growth experiments see growth-engine; for content optimization see autoresearch.

Maintain SESSION-STATE.md, the durable within-session memory that survives context compaction. Use when a constraint, settled decision, existing-infrastructure (survey) finding, or open thread must persist across a long session so it is not re-derived or re-litigated. Triggers on /state, "remember this for the session", "record this constraint/decision", or after surveying what already exists. For cross-session/personal memory use .claude/memory/ on Claude or native memories on Codex; for repo-derivable facts, do not record at all.

Use when reviewing or auditing a library of Claude Code skills, agents, slash commands, and workflows — frontmatter correctness, routing quality, tool allowlists, command arg-hints, workflow meta/phase coherence, cross-reference coherence, single-responsibility, file structure, and anti-pattern detection. Triggers on mentions of "review skills", "audit agents", "skill library", "agent definition review", "review this command", "review this workflow", "is this skill right", or when iterating on `.claude/skills/`, `.claude/agents/`, `.claude/commands/`, or `.claude/workflows/` directories. For code review of source code see code-review-and-quality.

Internal utility for the skills library itself — opt-in, local-first, privacy-respecting usage telemetry and update checks. Provides version checking, usage logging, and usage reporting for the skills in this repo. Use only when asked to view skill usage stats, configure telemetry opt-in/out, wire a skill to log its own runs, or check for skill updates.

Use when implementing analytics with PostHog in a TypeScript app — capturing events, identifying users, adding feature flags, tracking errors, or wiring API lifecycle telemetry. Triggers on edits to analytics integration files, or mentions of "PostHog", "analytics", "feature flag", "event tracking", "capture", "identify", "A/B test", or "experiment".

Use when building data pipelines, ETL jobs, event processors, message-broker producers/consumers, application caching layers, database migrations, BigQuery queries, or event-sourcing handlers in TypeScript. Triggers on edits to indexer/ETL/migration code, Prisma or Drizzle schemas, BigQuery integration code, RabbitMQ/Kafka/SQS/BullMQ producer or consumer code, Redis cache wrappers, or mentions of "data engineering", "ETL", "pipeline", "indexer", "event sourcing", "data warehouse", "data migration", "message queue", "RabbitMQ", "Kafka", "SQS", "BullMQ", "Redis cache", or "caching".

Use when writing or reviewing TypeScript backend tests — Jest unit tests for services/controllers (mocked Prisma) or Supertest integration tests against a real isolated PostgreSQL test database. Triggers on edits to `*.service.test.ts`, `*.controller.test.ts`, `*.integration.test.ts`, service/controller/API test files (`.ts`, not `.tsx`) under `**/__tests__/`, or mentions of "backend test", "service test", "API test", "database test".

Use when writing or reviewing TypeScript frontend tests — Jest unit/integration tests for React components and hooks built with Chakra UI, React Query, Zustand, and Next.js App Router. Triggers on edits to `*.test.tsx`, React component/hook test files (`.tsx`, not `.ts`) under `**/__tests__/`, custom test render helpers, or mentions of "frontend test", "component test", "hook test", "React test", "UI test".

Use when writing, reviewing, or deploying Solidity smart contracts — token contracts, signature-gated claim systems, merkle-gated staking vaults, allocation modules, role-based proxies, or factory patterns. Triggers on edits to `*.sol` files, `hardhat.config.*`, `foundry.toml`, deploy scripts, or mentions of "smart contract", "Solidity", "Web3", "EVM", "Hardhat", "Foundry", "ERC20/721/1155", "merkle", "staking", or "on-chain". For security audits and adversarial review see security-engineering.