Leonardo Gurgel
Community@leogurja · Teresópolis -RJ, Brasil
Frontend Developer | React | Next | TypeScript
Agent Skills by Leonardo Gurgel
Showing 38 vetted skills indexed across 1 GitHub repositories.
leogurjausing-git-worktrees
Creates isolated git worktree workspaces with native tool detection and baseline test verification.
leogurjacaveman-help
Displays a reference card of caveman modes, skills, and deactivation commands.
leogurjacaveman-evidence-review
Reviews Caveman Cloud cost, trace, and savings evidence through read-only queries.
leogurjainvestigate-first
Diagnose ambiguous failures by ranking evidence-based hypotheses before editing code.
leogurjatest-driven-development
Enforces red-green-refactor test-driven development workflow for features and bugfixes.
leogurjacaveman-explore
Locates relevant code in repositories and returns path:line citations via parallel read-only searches.
leogurjacaveman-learn
Applies consent-gated fixes to ranked token sinks from a Caveman learn report.
leogurjasystematic-debugging
Diagnose bugs through a four-phase root cause investigation process before proposing fixes.
leogurjacaveman-review
Generates one-line code review comments with line numbers, severity, and concrete fixes.
leogurjacaveman-compress
Compress natural language memory files into shorter caveman format to reduce input tokens.
leogurjausing-superpowers
Enforces skill discovery and invocation before any AI response or action.
leogurjadispatching-parallel-agents
Dispatch concurrent subagents to fix independent test failures across separate problem domains.
leogurjacaveman-manage
Inspect Caveman Cloud experiment lifecycle state and block unsafe lifecycle mutations.
leogurjafind-skills
Searches and installs agent skills from the open skills ecosystem using the Skills CLI.
leogurjacaveman-setup
Wire repository LLM callsites through the Caveman gateway for spend measurement.
leogurjaexecuting-plans
Executes written implementation plans task-by-task with review checkpoints and verification steps.
leogurjasurgical-patch
Fix bugs by changing the narrowest responsible layer with regression proof.
leogurjafinishing-a-development-branch
Guides merging, pushing, or preserving completed git branches with worktree cleanup.
leogurjacaveman-optimize
Evaluates Caveman report-only optimization observations with operator-chosen candidates and paired baseline evaluations.
leogurjacaveman-discover
Find and label LLM workflows in a repository to break down gateway spend by workflow.
leogurjasafe-refactor
Restructures code while preserving behavior through verification-bracketed structural edits.
leogurjaverify-and-stop
Validates acceptance criteria with focused checks and stops without expanding scope.
leogurjaorca-cli
Operate Orca worktrees, terminals, repos, and artifacts through the orca CLI.
leogurjacavecrew
Decide when to delegate code tasks to compressed-output subagents instead of working inline.
Frequently Asked Questions About Leonardo Gurgel
FAQPage SchemaWhat tasks can I accomplish with leogurja's skills?▼
The 38 skills cover test-driven development, systematic debugging, safe refactoring, surgical patching, reversible migrations, plan writing and execution, code review request and response, git worktree isolation, parallel agent dispatch, LLM spend discovery and optimization, memory-file compression, and Conventional Commits message writing.
Who are these skills designed for?▼
They target software engineers and agent operators who want disciplined development practices, such as evidence-ranked debugging and verification-before-completion, plus teams running LLM agents who need token cost visibility, compressed context, and supervised multi-worker coordination through Orca and Caveman Cloud.
How do the Caveman skills reduce LLM costs in practice?▼
caveman-setup routes repository LLM requests through the Caveman Cloud gateway for measurement, caveman-discover labels workflows so spend is grouped per workflow, caveman-learn applies ranked token-saving fixes with consent, and caveman/caveman-compress shrink output tokens and memory files like CLAUDE.md.
What prerequisites or dependencies do these skills require?▼
Most discipline skills (TDD, debugging, refactoring, verification) need only a repository and an agent session. Orca-related skills require the `orca` CLI and Orca-managed worktrees; Caveman skills require a Caveman Cloud gateway connection; caveman-explore uses Read, Glob, and Grep with the haiku model.
How do the skills enforce code quality before merging?▼
verification-before-completion requires running verification commands and confirming output before claiming success, test-driven-development mandates tests before implementation code, requesting-code-review and caveman-review produce compressed findings with location and fix, and finishing-a-development-branch guides integration once all tests pass.