seasonedcc
- 139 skills
- 0 followers
- 5 hours ago last updated
- ▌ Quick 2 · seasonedccRun a small fix or polish to an existing surface under a reduced quick-iteration Definition of Done — no subagents, no mobile pass, no self-improvement pass. Use ONLY when the user explicitly invokes /quick; never select this mode on your own, and never suggest it.
- ▌ Kysely 2 · seasonedccWrite Kysely queries and migrations following project conventions. Use when writing migrations, creating tables, adding columns, creating enums, writing database queries with Kysely, or using database transactions.
- ▌ Kysely 3 · seasonedccWrite Kysely queries and migrations following project conventions. Use when writing migrations, creating tables, adding columns, creating enums, writing database queries with Kysely, or using database transactions.
- ▌ Release 2 · seasonedccCut a production release — pick the next v<N> tag, deeply study everything merged since the last release, write curated release notes grouped by shaping project, solved meeting requests, and topic, stamp solved requests-from-meetings entries with Released badges, audit everything through a pre-release PR, then publish the GitHub Release that triggers the deploy. Use when the user invokes /release, asks to cut a release, release to production, ship what's on main, or write release notes.
- ▌ Release 3 · seasonedccRelease new versions of the composable-controllers packages via Changesets. Analyze the diff, write the changeset, run checks, commit, ask the user to publish, and create per-package GitHub releases. Use when the user mentions releasing, publishing, cutting a release, or shipping a new version.
- ▌ Testing 2 · seasonedcc bundleWrite and run tests using Vitest (unit tests) and Playwright (E2E tests). Use when writing tests, fixing tests, running tests, implementing TDD, or when user mentions testing, test-driven development, Vitest, Playwright, unit tests, or E2E tests.
- ▌ Testing 3 · seasonedccWrite and run tests for product-monolith using pytest (unit tests) and Playwright via pytest-playwright (E2E tests). Use when writing tests, fixing tests, running tests, implementing TDD, seeding E2E data, or when the user mentions testing, test-driven development, unit tests, E2E tests, Playwright, or mutation proofs.
- ▌ Env Vars 2 · seasonedccManage environment variables following the two-tier env pattern. Use when adding new env vars, modifying env schemas, working with app/framework/env.server.ts or app/env.server.ts, or updating CI env configuration.
- ▌ Quick 3 · seasonedccRun a small fix or polish to an existing surface under a reduced quick-iteration Definition of Done — no subagents, no self-improvement pass. Use ONLY when the user explicitly invokes /quick; never select this mode on your own, never suggest it, and never let it self-trigger.
- ▌ Quick 4 · seasonedccRun a small fix or polish to an existing surface under a reduced quick-iteration Definition of Done — no subagents, no self-improvement pass. Use ONLY when the user explicitly invokes /quick; never select this mode on your own, never suggest it, and never let it self-trigger.
- ▌ Kysely 4 · seasonedccWrite Kysely queries and migrations following project conventions. Use when writing migrations, creating tables, adding columns, creating enums, writing database queries with Kysely, or using database transactions.
- ▌ Main Sync 2 · seasonedccSync a long-lived feature branch with its repo's advancing main — merge main in, judge the rule-set and capability deltas, and re-run the combined Definition of Done. Use when main gains commits under a goal or review-fixes feature branch, before merging origin/main into a feature branch, or when deciding whether a feature landed on main must be mirrored on the branch.
- ▌ Pr Review 2 · seasonedccReview a GitHub pull request with full context — the description, every comment, review, and inline thread, then the diff — judging the big picture personally before multi-angle finding and adversarial verification. Use when asked to review a PR, analyze a pull request, or assess a teammate's changes — including a PR by the founder, whose PRs the team takes over; use this instead of the native /review command.
- ▌ Subagents 2 · seasonedccSpawn subagents and dynamic workflows well — size each task to the context window (~33% of 1M target) and pick the right model tier. Use whenever delegating work to subagents, launching a Workflow, or deciding how to split a task across agents.
- ▌ Subagents 3 · seasonedccSpawn subagents and dynamic workflows well — size each task to the context window (~33% of 1M target), pick the right model tier, and rule on agents that run past it. Use whenever delegating work to subagents, launching a Workflow, deciding how to split a task across agents, or handling a watchdog alert about a full context.
- ▌ Subagents 4 · seasonedccSpawn subagents and dynamic workflows well — size each task to the context window (~33% of 1M target), pick the right model tier, and rule on agents that run past it. Use whenever delegating work to subagents, launching a Workflow, deciding how to split a task across agents, or handling a watchdog alert about a full context.
- ▌ Worktrees 2 · seasonedccRun parallel tasks in isolated git worktrees, each with its own databases, ports, env files, and dev server. Use when creating or removing worktrees, running parallel tasks or agents, starting a dev server for a specific worktree, running browser tests against an isolated server, or when the user mentions worktree, parallel lanes, or isolated environments.
- ▌ Worktrees 3 · seasonedccRun parallel tasks in isolated git worktrees, each with its own databases, ports, env files, and dev server. Use when creating or removing worktrees, running parallel tasks or agents, starting a dev server for a specific worktree, running browser tests against an isolated server, or when the user mentions worktree, parallel lanes, or isolated environments.
- ▌ Worktrees 4 · seasonedcc bundleRun parallel tasks in isolated git worktrees, one per workspace repo being changed, provisioned by the setup/teardown scripts (env files, databases, ports for product-monolith). Use when creating or removing worktrees, running parallel tasks or agents, starting a dev server for a specific lane, or when the user mentions worktree, parallel lanes, or isolated environments.
- ▌ MCP Server 2 · seasonedccExtend and maintain the app's MCP server — add tools for new or existing app capabilities, keep the CI parity check green, and work with the OAuth 2.1 layer. Use when adding or changing any business function or route (the parity test will demand MCP coverage), when writing MCP tools, when working on app/mcp/, the /mcp endpoint, OAuth routes, scopes, or the consent screen, or when the user mentions MCP, tools, or the parity check.
- ▌ MCP Server 3 · seasonedccBuild and extend the app MCP server — the machine surface that exposes app capabilities as tools over OAuth 2.1. Use when adding or changing an MCP tool, wiring a new domain into the MCP server, touching app/mcp/*, the oauth business layer, or the /mcp, /oauth/*, or /.well-known/* routes, when working on the parity check or pending-coverage, or when the user mentions MCP, tools, Model Context Protocol, OAuth, bearer tokens, scopes, or the consent screen.
- ▌ Release 4 · seasonedccCut a release — scope the span since the last release, deeply study every merged pull request in it, write release notes in which every sentence traces to a merged change, run the gates and audits, and treat the publish as the deliberate final act. Use when the user asks to release, cut a release, publish, ship a new version, ship what's on main, release to production, or write release notes.
- ▌ Shaping 2 · seasonedccRun framing and shaping sessions that end in a shaping document, and answer questions about the method from the verbatim books and posts in references/. Use when framing a raw idea, shaping a project, writing or revising a shaping document under shaping/, working feedback on one back in, breadboarding or drawing fat marker sketches, finding and defusing time bombs (Shape Up's rabbit holes), naming no-gos, or when reasoning about customer demand — jobs to be done, JTBD, the struggling moment, forces of progress, demand-side selling, why people switch. Also the corpus to consult on anything Shape Up covers beyond our process — pitches, appetite, betting tables, scopes, hill charts, scope hammering, deciding when to stop.
- ▌ Testing 4 · seasonedccWrite, run, and debug tests across the project's unit and end-to-end suites. Covers test-driven development and mutation proofs, gate reconciliation, parallel-safe unit tests, E2E seeding and retry safety, coverage discipline, and diagnosing CI-only failures and flakes. Use when writing tests, fixing or diagnosing failing or flaky tests, running test suites, seeding E2E data, or when the user mentions testing, test-driven development, unit tests, E2E tests, or mutation proofs.
- ▌ Release 5 · seasonedccCut a release — scope the span since the last release, deeply study every merged pull request in it, write release notes in which every sentence traces to a merged change, run the gates and audits, and treat the publish as the deliberate final act. Use when the user asks to release, cut a release, publish, ship a new version, ship what's on main, release to production, or write release notes.
- ▌ Testing 5 · seasonedcc bundleWrite, run, and debug tests across the project's unit and end-to-end suites. Covers test-driven development and mutation proofs, gate reconciliation, parallel-safe unit tests, E2E seeding and retry safety, coverage discipline, and diagnosing CI-only failures and flakes. Use when writing tests, fixing or diagnosing failing or flaky tests, running test suites, seeding E2E data, or when the user mentions testing, test-driven development, unit tests, E2E tests, or mutation proofs.
- ▌ Type Safety 2 · seasonedccWrite minimal, correct TypeScript type annotations. Use when adding types, declaring variables, writing function signatures, creating type aliases, or reviewing code for unnecessary type declarations.
- ▌ Reground 2 · seasonedccReground after a compaction — from this session's ledger and the real sources of truth, never from the compacted summary alone. Use ONLY when the user explicitly invokes /reground — the person types this command, never you.
- ▌ Quick 5 · seasonedccRun a small fix or polish to an existing surface under a reduced quick-iteration Definition of Done — no subagents, no self-improvement pass. Use ONLY when the user explicitly invokes /quick; never select this mode on your own, never suggest it, and never let it self-trigger.
- ▌ Env Vars 3 · seasonedccManage environment variables following the two-tier env pattern. Use when adding new env vars, modifying env schemas, working with app/framework/env.server.ts or app/env.server.ts, or updating CI env configuration.
- ▌ Demo Videos 2 · seasonedccProduce the two narrated demo videos for a scope of work — a five-minute highlights cut and a full-length walkthrough — filmed from a screenplay against the seeded product. Use when the user invokes /demo-videos, asks for a demo video, demo videos, a screencast, a product walkthrough video, a narrated demo, video narration, a recording of what shipped, or a video to show stakeholders what a project delivered.
- ▌ Post Review 2 · seasonedccPost a completed code review on a teammate's GitHub PR as inline comments anchored to the diff, with the verdict and blocking rationale in the review summary. Use when posting review findings to GitHub, requesting changes or approving a PR, taking over a PR by the founder, or after analyzing a PR (the pr-review skill covers the analysis) when the findings should land on the PR itself.
- ▌ Main Sync 3 · seasonedccSync a long-lived feature branch with its repo's advancing main — bring main's commits into the branch under the repository's sync strategy, judge the rule-set and capability deltas, and re-run the combined Definition of Done. Use when main gains commits under a goal or review-fixes feature branch, before syncing a feature branch with origin/main, or when deciding whether a feature landed on main must be mirrored on the branch.
- ▌ Pr Review 3 · seasonedccReview a GitHub pull request with full context — the description, every comment, review, and inline thread, then the diff — judging the big picture personally before multi-angle finding and adversarial verification. Use when asked to review a PR, analyze a pull request, assess a teammate's changes, or run the pre-merge adversarial review of a draft PR; use this instead of the native /review command.
- ▌ Subagents 5 · seasonedccSpawn subagents and dynamic workflows well — size each task to the context window (~33% of 1M target), pick the right model tier, and rule on agents that run past it. Use whenever delegating work to subagents, launching a Workflow, deciding how to split a task across agents, or handling a watchdog alert about a full context.
- ▌ Worktrees 5 · seasonedccRun parallel tasks in isolated git worktrees, each provisioned as a lane with its own databases, ports, env files, and services through the seasoned-skills binary. Use when creating or removing worktrees, provisioning or tearing down lanes, running parallel tasks or agents, starting a dev server for a specific lane, handing a lane to the user for manual testing, or when the user mentions worktree, parallel lanes, or isolated environments.
- ▌ Review Fixes 2 · seasonedccTurn a completed PR review's findings into an adjudicated fix scope and land the fixes on the reviewed PR's own branch — triage the certain defects, adjudicate everything else with the user one finding at a time, then orchestrate fix lanes merged into the feature branch. Use when fixing review findings on our own PR, addressing a pr-review analysis, or driving a reviewed PR toward a clean re-review.
- ▌ Authorization 2 · seasonedccImplement authorization using context getters, context schemas, and composable functions. Use when working with authentication, permissions, user access control, protected routes, business logic authorization, or when user mentions auth, context, permissions, or access control.
- ▌ Authorization 3 · seasonedccImplement authorization using context getters, context schemas, and composable functions. Use when working with authentication, permissions, user access control, protected routes, business logic authorization, or when user mentions auth, context, permissions, or access control.
- ▌ Nested Routes 2 · seasonedccCreate nested route sections with layouts, index redirects, and menu navigation in React Router v7. Use when adding new route groups, creating layouts with sidebar navigation, nesting routes under a parent path, or restructuring flat routes into nested ones.
- ▌ Optimistic UI 2 · seasonedccBuild optimistic UI with React Router using deterministic resource routes. Use when implementing toggles, optimistic updates, action-only routes, or when user mentions optimistic UI, toggle, or real-time feedback.
- ▌ Optimistic UI 3 · seasonedccBuild optimistic UI with React Router v7 using deterministic resource routes. Use when implementing toggles, optimistic updates, action-only routes, or when user mentions optimistic UI, toggle, or real-time feedback.
- ▌ Kysely 5 · seasonedccWrite Kysely queries and migrations following project conventions. Use when writing migrations, creating tables, adding columns, creating enums, writing database queries with Kysely, or using database transactions.
- ▌ Agent Browser 2 · seasonedcc bundleAutomate browser interactions using the agent-browser CLI. Use when navigating pages, clicking elements, filling forms, taking screenshots, extracting page data, running browser tests, or verifying UI behavior. Use when the user mentions browser, agent-browser, test in browser, navigate, click, screenshot, interact with page, or automate.
- ▌ Agent Browser 3 · seasonedcc bundleAutomate browser interactions using the agent-browser CLI, and keep its sessions and processes from leaking. Use when navigating pages, clicking elements, filling forms, taking screenshots, extracting page data, running browser tests, verifying UI behavior, or cleaning up leftover browser processes. Use when the user mentions browser, agent-browser, test in browser, navigate, click, screenshot, interact with page, automate, leaked or runaway browsers, Chrome eating memory, or killing processes.
- ▌ Agent Browser 4 · seasonedcc bundleAutomate browser interactions using the agent-browser CLI. Use when navigating pages, clicking elements, filling forms, taking screenshots, extracting page data, running browser tests, or verifying UI behavior. Use when the user mentions browser, agent-browser, test in browser, navigate, click, screenshot, interact with page, or automate.
- ▌ Design System 2 · seasonedccFollow the app's design system when building or changing UI. Use when creating pages, lists, forms, drawers, modals, badges, buttons, empty states, headings, or status indicators, when choosing icons or copy, or when the user mentions design system, UI patterns, visual consistency, typography, or UX canon.
- ▌ Design System 3 · seasonedccFollow the app design system when building or changing UI. Use when creating pages, lists, forms, drawers, modals, badges, buttons, empty states, headings, or status indicators, when choosing icons, copy, button sizes, or breakpoints, when making a surface work on phones, tablets, or paper, or when the user mentions design system, UI patterns, visual consistency, typography, responsiveness, mobile, printing, or UX canon.
- ▌ Orchestration 2 · seasonedccRun delegated work reliably — writing subagent charters, verifying subagent claims, keeping judgment at the orchestrator tier, ledger discipline, recovering from interruptions, and shipping/merging lane PRs. Use when spawning subagents or workflows, coordinating parallel lanes, resuming after a session limit or compaction, or rebasing and merging a lane's PR. The subagents skill covers task sizing and model choice; this skill covers everything after the spawn.
- ▌ Orchestration 3 · seasonedccRun delegated work reliably — writing subagent charters, verifying subagent claims, keeping judgment at the orchestrator tier, ledger discipline, recovering from interruptions, and shipping/merging lane PRs. Use when spawning subagents or workflows, coordinating parallel lanes, resuming after a session limit or compaction, or rebasing and merging a lane's PR. The subagents skill covers task sizing and model choice; this skill covers everything after the spawn.
- ▌ Orchestration 4 · seasonedccRun delegated work reliably — writing subagent charters, verifying subagent claims, keeping judgment at the orchestrator tier, ledger discipline, recovering from interruptions, and shipping/merging lane PRs. Use when spawning subagents or workflows, coordinating parallel lanes, resuming after a session limit or compaction, or rebasing and merging a lane's PR. The subagents skill covers task sizing and model choice; this skill covers everything after the spawn.
- ▌ Demo Videos 3 · seasonedccProduce the two narrated demo videos for a scope of work — a five-minute highlights cut and a full-length walkthrough — filmed from a screenplay against the seeded product. Use when the user invokes /demo-videos, asks for a demo video, demo videos, a screencast, a product walkthrough video, a narrated demo, video narration, a recording of what shipped, or a video to show stakeholders what a project delivered.
- ▌ Post Review 3 · seasonedccPost a completed code review on a teammate's GitHub PR as inline comments anchored to the diff, with the verdict and blocking rationale in the review summary. Use when posting review findings to GitHub, requesting changes or approving a PR, or after analyzing a PR (the pr-review skill covers the analysis) when the findings should land on the PR itself.
- ▌ Shaping 3 · seasonedccRun framing and shaping sessions that end in a shaping document, and answer questions about the method from the verbatim books and posts in references/. Use when framing a raw idea, shaping a project, writing or revising a shaping document under shaping/, working feedback on one back in, breadboarding or drawing fat marker sketches, finding and defusing time bombs (Shape Up's rabbit holes), naming no-gos, or when reasoning about customer demand — jobs to be done, JTBD, the struggling moment, forces of progress, demand-side selling, why people switch. Also the corpus to consult on anything Shape Up covers beyond our process — pitches, appetite, betting tables, scopes, hill charts, scope hammering, deciding when to stop.
- ▌ Testing 6 · seasonedcc bundleWrite, run, and debug tests across the project's unit and end-to-end suites. Covers test-driven development and mutation proofs, gate reconciliation, parallel-safe unit tests, E2E seeding and retry safety, coverage discipline, and diagnosing CI-only failures and flakes. Use when writing tests, fixing or diagnosing failing or flaky tests, running test suites, seeding E2E data, or when the user mentions testing, test-driven development, unit tests, E2E tests, or mutation proofs.
- ▌ Demo Videos 4 · seasonedccProduce the two narrated demo videos for a scope of work — a five-minute highlights cut and a full-length walkthrough — filmed from a screenplay against the seeded product. Use when the user invokes /demo-videos, asks for a demo video, demo videos, a screencast, a product walkthrough video, a narrated demo, video narration, a recording of what shipped, or a video to show stakeholders what a project delivered.
- ▌ Type Safety 3 · seasonedccWrite minimal, correct TypeScript type annotations. Use when adding types, declaring variables, writing function signatures, creating type aliases, or reviewing code for unnecessary type declarations.
- ▌ Background Jobs 2 · seasonedccCreate and manage background jobs with graphile-worker using makeJob and makeCronJob. Use when adding new background jobs, cron jobs, fan-out patterns, enqueuing work, or working with the worker infrastructure.
- ▌ Background Jobs 3 · seasonedccCreate and manage background jobs with graphile-worker using makeJob and makeCronJob. Use when adding new background jobs, cron jobs, fan-out patterns, enqueuing work, or working with the worker infrastructure.
- ▌ Database Design 2 · seasonedccDesign database tables and migrations following project conventions. Use when creating new tables, writing migrations, adding columns, or discussing database schema design.
- ▌ Database Design 3 · seasonedccDesign database tables and migrations following the project's zero-exception append-only, event-sourced doctrine. Use when creating new tables, writing migrations, adding columns, modeling state changes, deriving current state, or discussing database schema design.
- ▌ Review Fixes 3 · seasonedccTurn a completed PR review's findings into an adjudicated fix scope and land the fixes on the reviewed PR's own branch — triage the certain defects, adjudicate everything else with the user one finding at a time, then orchestrate fix lanes merged into the feature branch. Use when fixing review findings on our own PR, addressing a pr-review analysis, or driving a reviewed PR toward a clean re-review.
- ▌ Main Sync 4 · seasonedccSync a long-lived feature branch with its repo's advancing main — bring main's commits into the branch under the repository's sync strategy, judge the rule-set and capability deltas, and re-run the combined Definition of Done. Use when main gains commits under a goal or review-fixes feature branch, before syncing a feature branch with origin/main, or when deciding whether a feature landed on main must be mirrored on the branch.
- ▌ Framework Folder 2 · seasonedccEnforce separation between framework and app code in app/framework/. Use when adding new files or logic to app/framework/, importing framework modules, or deciding where to place new abstractions.
- ▌ Orchestration 5 · seasonedccRun delegated work reliably — writing subagent charters, verifying subagent claims, keeping judgment at the orchestrator tier, ledger discipline, recovering from interruptions, and shipping/merging lane PRs. Use when spawning subagents or workflows, coordinating parallel lanes, resuming after a session limit or compaction, or rebasing and merging a lane's PR. The subagents skill covers task sizing and model choice; this skill covers everything after the spawn.
- ▌ Agent Browser 5 · seasonedcc bundleAutomate browser interactions using the agent-browser CLI, and keep its sessions and processes from leaking. Use when navigating pages, clicking elements, filling forms, taking screenshots, extracting page data, running browser tests, verifying UI behavior, or cleaning up leftover browser processes. Use when the user mentions browser, agent-browser, test in browser, navigate, click, screenshot, interact with page, automate, leaked or runaway browsers, Chrome eating memory, or killing processes.
- ▌ Authorization 4 · seasonedccImplement authorization using context getters, context schemas, and composable functions. Use when working with authentication, permissions, user access control, protected routes, business logic authorization, or when user mentions auth, context, permissions, or access control.
- ▌ Design System 4 · seasonedccFollow the project's design system when building or changing UI. Use when creating pages, lists, forms, drawers, modals, badges, buttons, empty states, headings, or status indicators, when choosing icons, copy, button sizes, or breakpoints, when making a surface work on phones, tablets, or paper, or when the user mentions design system, UI patterns, visual consistency, typography, responsiveness, mobile, printing, or UX canon.
- ▌ Nested Routes 3 · seasonedccCreate nested route sections with layouts, index redirects, and menu navigation in React Router v7. Use when adding new route groups, creating layouts with sidebar navigation, nesting routes under a parent path, or restructuring flat routes into nested ones.
- ▌ Optimistic UI 4 · seasonedccBuild optimistic UI with React Router v7 using deterministic resource routes. Use when implementing toggles, optimistic updates, action-only routes, or when user mentions optimistic UI, toggle, or real-time feedback.
- ▌ Self Improvement 2 · seasonedccCodify lessons from a completed task into repo improvements — skill updates, CLAUDE.md changes, or root-cause tooling fixes — delivered as PRs the user reviews and merges. Use when a task has met every other Definition of Done criterion, when a /goal goal is met but not yet marked complete, or when the user asks to derive lessons from recent work.
- ▌ Self Improvement 3 · seasonedccCodify lessons from a completed task into repo improvements — skill updates, CLAUDE.md changes, or root-cause tooling fixes — delivered as PRs the user reviews and merges. Use when a task has met every other Definition of Done criterion, when a /goal goal is met but not yet marked complete, or when the user asks to derive lessons from recent work.
- ▌ Self Improvement 4 · seasonedccCodify lessons from a completed task into repo improvements — skill updates, CLAUDE.md changes, or root-cause tooling fixes — delivered as PRs the user reviews and merges. Use when a task has met every other Definition of Done criterion, when a /goal goal is met but not yet marked complete, or when the user asks to derive lessons from recent work.
- ▌ MCP Server 4 · seasonedccBuild and extend the app's MCP server — the machine surface that exposes app capabilities as tools over OAuth 2.1, held to capability parity with the product. Use when adding or changing an MCP tool, wiring a new domain into the MCP server, touching app/mcp/*, the oauth business layer, or the /mcp, /oauth/*, or /.well-known/* routes, when working on the parity check, the exception register, or pending-coverage, or when the user mentions MCP, tools, Model Context Protocol, OAuth, bearer tokens, scopes, or the consent screen.
- ▌ Seasoned Skills 2 · seasonedccInstall, configure, upgrade, and repair the seasoned-skills workflow package — the source of this project's generated doctrine, skills, and runtime code. Use when installing or upgrading the workflow, changing seasoned-skills.config.ts, running or debugging the sync, reading a failed-sync report, or wiring a new machine.
- ▌ Type Safety 4 · seasonedccWrite minimal, correct TypeScript type annotations. Use when adding types, declaring variables, writing function signatures, creating type aliases, or reviewing code for unnecessary type declarations.
- ▌ Background Jobs 4 · seasonedccCreate and manage background jobs with graphile-worker using makeJob and makeCronJob. Use when adding new background jobs, cron jobs, fan-out patterns, enqueuing work, or working with the worker infrastructure.
- ▌ Business Folder 2 · seasonedccOrganize business logic in app/business/ by domain cohesion and framework independence. Use when creating files in app/business/, adding functions to business files, naming business modules, deciding where to place business logic, or choosing what to import in business code.
- ▌ Database Design 4 · seasonedccDesign database tables and migrations following the project's identity-and-event-table doctrine. Use when creating new tables, writing migrations, adding columns, modeling state changes, deriving current state, or discussing database schema design.
- ▌ Self Improvement 5 · seasonedccCodify lessons from a completed task on two channels at once — a draft pull request on the project's own content or tooling, plus a detailed anonymized issue on the seasoned-skills package — with every PR marked ready only by the user. Use when a task has met every other Definition of Done criterion, when a /goal goal is met but not yet marked complete, or when the user asks to derive lessons from recent work.
- ▌ Skill Management 2 · seasonedccCreate, manage, and debug Claude Code Agent Skills. Use when creating new skills, debugging skill activation issues, writing SKILL.md files, managing skill structure, or learning about Claude Code skills. Helps with personal skills, project skills, YAML frontmatter, descriptions, and troubleshooting.
- ▌ Orchestration 6 · seasonedccRun delegated work reliably — writing subagent charters, verifying subagent claims, keeping judgment at the orchestrator tier, ledger discipline, recovering from interruptions, and shipping/merging lane PRs. Use when spawning subagents or workflows, coordinating parallel lanes, resuming after a session limit or compaction, or rebasing and merging a lane's PR. The subagents skill covers task sizing and model choice; this skill covers everything after the spawn.
- ▌ Framework Folder 3 · seasonedccEnforce separation between framework and app code in the framework folder. Use when adding new files or logic to the framework folder, importing framework modules, or deciding where to place new abstractions.
- ▌ Composable Functions 2 · seasonedcc bundleWork with composable-functions library for type-safe business logic. Use when working with applySchema, withContext, pipe, sequence, business functions, context validation, input validation, or when user mentions composable functions, schemas, or error handling patterns.
- ▌ Composable Functions 3 · seasonedcc bundleWork with composable-functions library for type-safe business logic. Use when working with applySchema, withContext, pipe, sequence, business functions, context validation, input validation, or when user mentions composable functions, schemas, or error handling patterns.
- ▌ Formatting Datetimes 2 · seasonedccFormat dates and timestamps in SQL queries instead of JS Date instances. Use when rendering dates, date/times, timestamps, or times from the database, when user mentions date formatting, toLocaleDateString, toLocaleString, or timezone issues, or when working with SSR hydration mismatches involving dates.
- ▌ Orchestration 7 · seasonedccRun delegated work reliably — writing subagent charters, verifying subagent claims, keeping judgment at the orchestrator tier, ledger discipline, recovering from interruptions, and shipping/merging lane PRs. Use when spawning subagents or workflows, coordinating parallel lanes, resuming after a session limit or compaction, or rebasing and merging a lane's PR. The subagents skill covers task sizing and model choice; this skill covers everything after the spawn.
- ▌ Background Jobs 5 · seasonedccCreate and manage background jobs with graphile-worker using makeJob and makeCronJob. Use when adding new background jobs, cron jobs, fan-out patterns, enqueuing work, or working with the worker infrastructure.
- ▌ Database Design 5 · seasonedccDesign database tables and migrations following the project's identity-and-event-table doctrine. Use when creating new tables, writing migrations, adding columns, modeling state changes, deriving current state, or discussing database schema design.
- ▌ Requests From Meetings 2 · seasonedccParse meeting recordings into precise feature-request documents under ./requests-from-meetings. Use when the user invokes /requests-from-meetings, asks to parse a meeting or meeting backlog, asks what was requested or shown in a meeting, or mentions meeting recordings, transcripts, or feature requests from meetings.
- ▌ Skill Management 3 · seasonedccCreate, manage, and debug Claude Code Agent Skills. Use when creating new skills, debugging skill activation issues, writing SKILL.md files, managing skill structure, or learning about Claude Code skills. Helps with personal skills, project skills, YAML frontmatter, descriptions, and troubleshooting. Also use when someone asks to "wire up a helper skill" or where the operational playbooks live.
- ▌ Composable Functions 4 · seasonedcc bundleWork with composable-functions library for type-safe business logic. Use when working with applySchema, withContext, pipe, sequence, business functions, context validation, input validation, or when user mentions composable functions, schemas, or error handling patterns.
- ▌ Formatting Datetimes 3 · seasonedccFormat dates and timestamps in SQL queries instead of JS Date instances. Use when rendering dates, date/times, timestamps, or times from the database, when user mentions date formatting, toLocaleDateString, toLocaleString, or timezone issues, or when working with SSR hydration mismatches involving dates.
- ▌ Prepare For Compaction 2 · seasonedccMake this session's ledger current so its context can be compacted safely, then answer plainly that compacting is safe. Use ONLY when the user explicitly invokes /prepare-for-compaction — the person types this command, never you.
- ▌ Requests From Meetings 3 · seasonedccParse meeting recordings into precise feature-request documents under ./requests-from-meetings. Use when the user invokes /requests-from-meetings, asks to parse a meeting or meeting backlog, asks what was requested or shown in a meeting, or mentions meeting recordings, transcripts, or feature requests from meetings.
- ▌ Requests From Meetings 4 · seasonedccParse meeting recordings into precise feature-request documents under ./requests-from-meetings. Use when the user invokes /requests-from-meetings, asks to parse a meeting or meeting backlog, asks what was requested or shown in a meeting, or mentions meeting recordings, transcripts, or feature requests from meetings.
- ▌ Formatting Datetimes 4 · seasonedccFormat dates and timestamps in SQL queries instead of JS Date instances. Use when rendering dates, date/times, timestamps, or times from the database, when user mentions date formatting, toLocaleDateString, toLocaleString, or timezone issues, or when working with SSR hydration mismatches involving dates.
- ▌ Make Typed Env Usage · seasonedccGenerate code examples and usage patterns for the make-typed-env library. Use when writing examples, documentation, demos, answering questions about the make-typed-env API, or helping users set up type-safe environment variables with Standard Schema, Zod, Valibot, ArkType, or string-ts key transformations.
- ▌ Make Typed Storage Usage · seasonedccGenerate code examples and usage patterns for the make-typed-storage library. Use when working with typed storage, importing from "make-typed-storage", or asking about type-safe Web Storage, makeTypedStorage, makeLocalStorage, makeSessionStorage, StorageLike, TypedStorage, StorageValidationError.
- ▌ Safe Element Comparison · seasonedccMaintain element type comparison safety in remix-forms. Use when modifying component mappings, adding new components to mapChildren, investigating unexpected HTML attributes or prop injection, writing docs about component customization, or responding to user issues about wrong elements receiving props.