Agent Skills by babidontcryy
Showing 64 vetted skills indexed across 1 GitHub repositories.
git-lessons
Extract actionable lessons from git fix commits by analyzing diffs and patterns.
verify-before-done
Run verification commands and attach outputs before asserting task completion.
worktree-safety
Create a feature branch and run baseline tests before code changes.
debug-protocol
Require a written root-cause and falsification test before code changes.
branch-finish
Validate tests, diffs, and commits before merging or discarding branches.
karpathy-guidelines
Enforce behavioral guidelines for writing, reviewing, and refactoring code.
tdd
Automate RED→GREEN→REFACTOR cycles with multi-agent context isolation.
plan-first
Draft a coding plan requiring user approval before execution.
gitnexus-refactoring
Automate code refactoring by mapping dependencies and updating imports across a codebase.
gitnexus-debugging
Trace error symptoms to execution paths and call graphs using GitNexus queries.
gitnexus-exploring
Trace execution flows and architectural relationships in large codebases using GitNexus.
gitnexus-guide
Answer questions about GitNexus tools, resources, and graph schema references.
gitnexus-cli
Run GitNexus CLI commands to analyze and index code repository knowledge graphs.
gitnexus-pr-review
Map PR diffs to execution flows and assess blast radius.
gitnexus-impact-analysis
Map direct and indirect dependents of code changes and assess risk.
research
Route research questions to specialist skills or fallback workflows with transparent decisions.
llm-cost-optimizer
Reduce LLM API spend with model routing, caching, and prompt compression.
agent-introspection-debugging
Capture failure state, diagnose root causes, and generate introspection reports for AI agent runs.
audit-env-variables
Audit environment variables in JavaScript/TypeScript projects and generate a Markdown ENV_AUDIT report.
telemetry-analysis
Analyze local YAMTAM telemetry logs for hook activity, token usage, and trust health.
code-tour
Create JSON CodeTour walkthroughs anchored to specific files and lines.
lsp-navigation
Identify symbol definitions and references using LSP navigation.
hook-protect-secrets
Block secret file access and credential exfiltration via PreToolUse hooks.
subagent-dependency
Coordinate subagent dependencies for parallel execution and safe result merging.