David Innes
Community@noobygains
Work in IT for Hitachi Solutions LTD - London
Agent Skills by David Innes
Showing 36 vetted skills indexed across 1 GitHub repositories.
reference-engine
Route development tasks to established domain-specific reference implementations before writing code.
quality-enforcement
Enforce code quality standards across pre-commit, pre-merge, and CI pipeline stages.
parallel-execution
Dispatch one agent per isolated problem domain for concurrent investigation and implementation.
merge-protocol
Validate test suites and present merge, pull request, deferral, or abandonment options.
performance-tuning
Measure software performance before and after optimizing identified bottlenecks.
test-first
Enforce Test-First Development by requiring a failing test before implementation code.
team-orchestration
Coordinate AI agent teams with task assignment and git worktree isolation.
protocol-authoring
Create and refine AI agent protocols using test-driven development cycles.
design-research
Survey template marketplaces and extract design patterns for industry-aligned web layouts.
project-bootstrap
Automate foundational setup for new software projects with standardized structure and tooling.
codebase-research
Search existing codebases for patterns, conventions, and similar implementations before writing new code.
comprehension-check
Require structured plain-language walkthroughs of code changes before commit.
activation
Enforce mandatory skill invocation before agent responses or actions.
design-integration
Detect a project's design system and enforce its conventions for new UI components.
ui-engineering
Guide frontend component architecture with layout, state, and accessibility decisions.
error-recovery
Track consecutive AI agent failures and enforce escalation thresholds.
delegated-execution
Execute implementation plans by dispatching tasks to isolated subagents with two-stage reviews.
quality-gate
Dispatch a code-reviewer agent to inspect commit ranges for quality standards.
agent-messaging
Define protocols for dispatching subagents and composing team messages.
security-protocol
Enforce validation and sanitization of external data crossing trust boundaries.
intent-discovery
Guide feature design through requirement clarification and approach evaluation.
fault-diagnosis
Diagnoses root causes of bugs and test failures through structured four-phase analysis.
task-runner
Executes approved implementation plans in batches with pauses for feedback.
review-response
Validate code review feedback against codebase and project requirements.