carlitose/agent-skills

ticket-autopilot

Drive a ticket folder AFK through deterministic scheduling, isolated implementation, evidence-backed quality gates, provider-neutral PRs, and explicit merge authorization.

llm-wiki

Build and maintain a self-compiling markdown wiki — an Agent ingests raw sources and a project's own docs, compiles cross-linked concept/entity/source pages, builds a dated timeline from Git and session history, answers queries against the corpus, lints the…

code-review

Review a runner candidate or standalone PR, commit, local diff, or requested scope read-only for regressions, evidence gaps, and overclaims.

code-simplification

Simplify the current candidate diff for clarity while preserving behavior, scope, tests, errors, side effects, and project conventions.

execute-ticket

Implement one normalized ticket through a bounded quality loop and return a validated handoff without scheduler or Git finalization side effects.

qa-test-plan

Plan causal QA for a runner candidate or standalone PR, commit, local diff, or requested scope without executing or deciding release.

verification-audit

Produce, validate, and reduce the canonical Verification Record for a frozen candidate without fabricating evidence or authorization.

handoff

Create a small, redacted, pointer-based handoff in the operating-system temporary directory so a fresh session can continue non-ticket work. Use when the user explicitly asks for a handoff or session transfer.

improve-codebase-architecture

Explore a codebase to find opportunities for architectural improvement, focusing on making the codebase more testable by deepening shallow modules. Use when user wants to improve architecture, find refactoring opportunities, consolidate tightly-coupled…

wizard

Author a staged Bash wizard for a manual setup, credential, migration, or cutover procedure that a human will run explicitly. Use when the user asks for a repeatable guided procedure with human-only actions.

resolving-merge-conflicts

Trace both sides of an in-progress Git merge or rebase conflict, resolve compatible intent with scoped authority, and gate incompatible intent. Use only when the user explicitly asks to inspect or resolve current conflicts.

to-questionnaire

Draft a focused questionnaire for an explicit external decision owner when the current user cannot answer alone. Use only when the user asks to externalize questions; never send the draft.

writing-for-agents

Improve documents that agents consume, including skills, AGENTS.md, CLAUDE.md, prompts, and linked instructions. Use when creating or editing agent-facing guidance whose triggers, hierarchy, completion criteria, or wording need to be clearer.

tdd

Test-driven development with red-green-refactor loop. Use when user wants to build features or fix bugs using TDD, mentions "red-green-refactor", wants integration tests, or asks for test-first development.

codebase-design

Shared vocabulary for designing deep, testable modules and evaluating interfaces, implementations, seams, adapters, leverage, and locality. Use when shaping a module boundary, comparing interface designs, classifying dependencies, or deciding where a test…

diagnose

Diagnose one bug by reproducing it, testing hypotheses, locating the root cause, and returning an evidence-backed diagnosis report. Use when the user wants root-cause analysis before implementation, or when triangulate-diagnosis delegates an independent…

triangulate-diagnosis

Diagnose a hard bug from three independent angles at once, then record the consensus as a diagnostic or decision spec and create follow-up tickets when needed. Use when a bug is high-stakes, has resisted a single diagnosis pass, or when the user wants to…

research

Research a factual, codebase, product, or external-documentation question using primary sources and return an evidence-backed answer. Use when the user asks to investigate, compare docs, find current behavior, answer a research ticket, or gather facts before…

to-spec

Create or update a focused feature, decision, diagnostic, architecture, or bug-analysis spec; backward compatibility is opt-in.

to-tickets

Break a spec into independently-grabbable tracer-bullet tickets and emit each versioned Ticket Envelope through the canonical scheduler contract.

wayfinder

Maintain a persistent map and investigation frontier for huge or vague work, using canonical normalized tickets and opt-in compatibility.

codebase-improver

End-to-end, self-contained Python/TypeScript workflow that maps and audits a whole repo, then recursively deepens modules. Use to improve or modernize a codebase, reduce technical debt, make this more testable, find refactoring opportunities, deepen shallow…

grill-me

A relentless interview to sharpen a plan or design. Compatibility alias for grilling; use when the user says "grill me", wants to stress-test a plan, or asks for a design interview.

grill-with-docs

A relentless interview to sharpen a plan or design, while also creating or updating domain-modeling docs such as CONTEXT.md glossary entries and sparse ADRs when justified.

grilling

Grill the user relentlessly about a plan or design. Use when the user wants to stress-test a plan before building, asks for a design interview, or uses trigger phrases such as "grill me", "grill this", "stress test this plan", "poke holes", or "challenge this…

explain-pr

Render and optionally publish a deterministic PR body from a validated verification bundle without changing its claims.

ask-skills

Route a request to the smallest composable local skill flow without duplicating orchestration or removed super-autopilot behavior.

prototype

Build a throwaway prototype to answer one design, logic, data-model, state-machine, or UI question before committing to production code. Use when the user asks to prototype, sanity-check an approach, mock up a UI, explore design options, compare flows, or…

domain-modeling

Build and sharpen a project's domain model and ubiquitous language while planning or changing software. Use when the user wants glossary work, bounded-context language, domain terminology cleanup, ADR-worthy decisions, or docs that keep product language…

peer-programming

Use this skill whenever the user wants to peer program, pair program, or do collaborative coding where THE HUMAN writes the code and Claude only assists. Trigger on phrases like "pair program", "peer program", "let's code together", "I'll write you review",…

pr-antipattern-review

Review a diff (pull request, commit, or local changes) by comparing it against the project's BLUEPRINT.md slice catalog. Auto-detects the input format (git diff, GitHub/GitLab PR URL or number, commit SHA, pasted diff). Loads the blueprint for each project…

project-blueprint

Discover the architectural slices, conventions, and patterns of a codebase by examining its structure, then output a written BLUEPRINT.md per project under docs/architecture (plus a SYSTEM_MAP.md for monorepos) that captures what the project actually does.…