Tyler Evans
Community@tyevans · Portland, Or
Doing my best.
Agent Skills by Tyler Evans
Showing 54 vetted skills indexed across 1 GitHub repositories.
tackline
Orchestrate multi-step agentic workflows for research, analysis, and planning.
expand
Expand sparse findings into detailed descriptions with cited evidence.
gather
Collect structured findings from codebases, documentation, and configuration files.
discover
Match natural language goals to skills and composite pipelines from SKILL.md files.
do
Route natural language goals to appropriate skills and pipelines.
merge
Merge and deduplicate findings from conversation context, TaskList outputs, and SendMessage history.
verify
Verify claims against source code, documentation, and external resources.
plan
Sequence tasks into dependency-aware execution order with prerequisite prioritization.
distill
Condenses prior LLM outputs or conversations into bullets, a paragraph, or a topic filter.
assess
Categorizes input items into predefined rubric-based verdicts, emitting grouped results with bolded labels.
diff-ideas
Compare two technical approaches across dimensions and generate a recommendation.
filter
Filter collection items by applying binary keep or drop criteria.
critique
Review proposals and plans to identify flaws, gaps, and risks.
rank
Score and reorder items by user-specified criteria with tie-breaking rules.
decompose
Decompose large goals into 3-6 bounded sub-parts with scope boundaries.
transform
Apply a rewrite instruction to each item in a collection.
sketch
Generate structural skeletons with TODO placeholders for code, documents, and workflows.
tracer
Implement features iteratively via thin end-to-end paths with validation and tests.
consensus
Synthesize multi-agent perspectives on design problems into a structured recommendation report.
session-health
Diagnoses AI conversation sessions for context load, scope coherence, and quality degradation signals.
bootstrap
Automate Claude Code project setup with CLAUDE.md, hooks, rules, and agents.
optimize
Measure, profile, and refactor software components to reduce latency and resource usage.
storm-prep
Analyze codebases to discover domain events and output structured YAML.
diagnose-agent
Analyze agent performance from git logs and learnings files to generate a struggle profile.