Agent Skills by lshtram
Showing 20 vetted skills indexed across 2 GitHub repositories.
Systematic Debugging
Guide software debugging through observation, instrumentation, hypothesis testing, and fix verification.
accordo-presentations
Generate Slidev presentations from topics or documents with Mermaid diagrams.
research-mastery
Validate library APIs and versions against official docs and codify verified examples into TECH_SPEC.md.
test-architect
Design pre-code testing strategies covering unit, integration, and end-to-end tests.
env-security
Scan codebases for secrets and generate .env.example scaffolds with dummy values.
supabase-mastery
Enforce migration-based Supabase schema changes with RLS and auditing.
self-correction
Verify code drafts against requirements with up to three retry attempts.
knowledge-integration
Distill task learnings into structured knowledge integration proposals for master guidelines.
context-management
Automates token and memory organization with tiered context management.
pattern-enforcement
Enforce coding patterns from CODING_STYLE.md and COMPONENT_PATTERNS.md across projects.
start-task
Create isolated git worktree environments with a task.md template.
perspective-engineering
Simulate three specialized viewpoint debates to surface blind spots in complex decisions.
github-automation
Coordinate Git operations, PR creation, and GitHub Action configurations.
security-audit
Identify and remediate software vulnerabilities using OWASP guidelines and STRIDE threat modeling.
strong-prompt
Analyze user requests for clarity, safety, and strategic alignment.
doc-maintainer
Detect and fix drift between code and TECH_SPEC_current.md documentation.
spec-writer
Interview users to generate structured PRDs with traceability tables.
meta-prompting
Design optimal prompts and task breakdowns for complex AI reasoning.
sequential-thinking
Organize complex tasks into sequential, verifiable steps with dependency ordering.
finish-task
Verify code quality and merge feature branches into main.