tool-discovery
Recommend the right agents and skills for any task. Covers both heavyweight agents (Task tool) and lightweight skills (Skill tool). Triggers on: which agent, which skill, what tool should I use, help me choose, recommend agent, find the right tool.
Install with Tessl CLI
npx tessl i github:NeverSight/skills_feed --skill tool-discovery95
Does it follow best practices?
Validation for skill structure
Evaluation results
100%
66%AI Assistant Onboarding Guide
Skill vs Agent decision logic
Skills for reference/lookup
75%
100%
Agents for reasoning/decisions
87%
100%
file-search for file finding
0%
100%
sql-patterns for CTE
12%
100%
postgres-expert for optimization
37%
100%
data-processing for YAML
12%
100%
rest-patterns for HTTP status
12%
100%
python-env for Python setup
12%
100%
Plan for architecture
37%
100%
git-workflow for git tasks
12%
100%
Skill invocation syntax
0%
100%
Agent invocation syntax
0%
100%
Max 2-3 tools per task
100%
100%
Without context: $0.1787 · 1m 11s · 9 turns · 14 in / 3,187 out tokens
With context: $0.1891 · 44s · 12 turns · 14 in / 2,426 out tokens
100%
90%Developer Workflow Automation Script
Correct tool invocation syntax
find-replace for batch rename
0%
100%
typescript-expert for TS generics
0%
100%
structural-search for AST
0%
100%
mcp-patterns for MCP protocol
0%
100%
react-expert for React architecture
0%
100%
Skill tool syntax for skills
0%
100%
Task tool syntax for agents
0%
100%
Skill-first principle stated
0%
100%
No task exceeds 3 tools
100%
100%
Without context: $0.1583 · 48s · 11 turns · 18 in / 2,545 out tokens
With context: $0.2343 · 52s · 14 turns · 699 in / 2,809 out tokens
100%
87%AI Tool Selection Reference Card
Specialist agent and catalog lookup
cloudflare-expert for Workers/KV/R2
0%
100%
sqlite-ops for SQLite async
0%
100%
tailwind-patterns for Tailwind
0%
100%
Plan agent for system design
0%
100%
structural-search for AST search
0%
100%
python-expert for pytest/async
0%
100%
Skill invocation syntax correct
0%
100%
Agent invocation syntax correct
0%
100%
Two-category principle stated
44%
100%
Catalog reference mentioned
11%
100%
No task exceeds 3 tools
100%
100%
Without context: $0.2729 · 1m 39s · 13 turns · 19 in / 5,040 out tokens
With context: $0.2034 · 49s · 11 turns · 14 in / 2,289 out tokens
- Evaluated
- Agent
- Claude Code
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.