Skip to the rows

agentsclimarketplace

Skill orchestrator

Skill HeyEddi-com/skills/evals/runs/2026-07-04T22-58-59Z/skill-orchestrator-suggest/.agents/skills/skill-orchestrator

HeyEddi Cursor Agent Skills hub — 20 skills, agent eval suite, git subtree distribution

Install

npx -y skills add HeyEddi-com/skills --skill skill-orchestrator

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Discover HeyEddi skills, load the catalog, and suggest which @skills to invoke for the current task. Use at session start, ambiguous requests, multi-skill pipelines, or when the user asks what skills are available. Reads skill-routing.json from @heyeddi-intake when present.

SKILL.md

2.6 KB, 611 tokens by cl100k_base, as published. Nobody here has run it

Skill Orchestrator

Routes work to the right HeyEddi skill — without loading every SKILL.md at once.

When to use

Session start on a HeyEddi project (.heyeddi/ present or greenfield app request)
User asks "what skills do we have?" or "which skill should handle this?"
Before a multi-step pipeline (intake → scaffold → design → handoff → QA)
After @heyeddi-intake — confirm skill-routing.json before downstream work

Limits (Cursor)

Cursor cannot inject all skill bodies automatically. This skill:

write_skills_index — scan once → .heyeddi/skills-index.json + skills-index.md
load_catalog — read the cached index (rescan with --refresh after skill installs)
suggest_skills — ranked matches from index entries (+ skill-routing.json when present)
You — read one chosen skill's SKILL.md, then invoke it

Pipeline

write_skills_index        → .heyeddi/skills-index.{json,md}  (session start / after installs)
read skills-index.md      → quick catalog without opening every SKILL.md
suggest_skills            → ranked @skills for this task
read one SKILL.md         → follow that skill's pipeline

If .heyeddi/docs/intake/skill-routing.json exists (from @heyeddi-intake), follow route order — suggestions surface it at score 100.

Instructions

Session start: write_skills_index --project-root <root> (or load_catalog --refresh — same effect)
Quick lookup: read .heyeddi/skills-index.md — no need to open every SKILL.md
suggest_skills --project-root <root> --user-prompt "<request>"
Or --prompt-file USER_PROMPT.md
Read one chosen skill's SKILL.md (path in index JSON), then invoke @skill-name
Re-run write_skills_index after npx skills add or hub skill updates
Greenfield with no .heyeddi/product.md → top suggestion often @heyeddi-intake

Tools

Script	Purpose
`write_skills_index.py`	Scan → `.heyeddi/skills-index.*`
`load_catalog.py`	Read index (or scan+write if missing)
`suggest_skills.py`	Rank skills from cached index

Related

@heyeddi-intake — writes skill-routing.json
reference/skill-triggers.md — how per-skill reference/triggers.md works (fully agnostic scoring)

What ships with it: 11 files

26.7 KB alongside SKILL.md, 5 of them executable

context/

reference/

skill-triggers.md1.3 KB
triggers.md111 B

scripts/

_catalog.pyruns17.5 KB
load_catalog.pyruns1.2 KB
_skill_cli.pyruns593 B
suggest_skills.pyruns1.6 KB
write_skills_index.pyruns1.2 KB

manifest.json1.7 KB

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.