v0.2.0 — Claude Code Plugin

Stop guessing which skills work.

Community-powered skill intelligence for Claude Code. Track outcomes, discover what works, zero PII.

claude plugins marketplace add sliday/claude-plugins
claude plugins install alit

Choosing skills is guesswork

200+ skills available. No data on which ones actually succeed. You pick a skill, it fails, you try another. Repeat. Skill authors have no feedback loop. Users have no signal.

Before

/find-skill "create PDF"

1. pdf          — no data
2. docx         — no data
3. document     — no data
4. latex        — no data

After (with alit)

/find-skill "create PDF"

1. pdf     87% success  1,423 uses
2. docx    62% success    891 uses
3. latex   41% success    204 uses
4. md-pdf  23% success     47 uses

Track. Report. Discover.

automatic

Track

After a skill runs, alit records the outcome. Three signals: heuristic output check, LLM self-assessment, user feedback.

anonymous

Report

Sends one anonymised telemetry row to the shared hive. No user IDs. No filenames. No correlation possible.

ranked

Discover

find-skill merges local registry with community data. Ranks by relevance (40%), success rate (40%), usage (20%).

Architecture

Module	Role
`registry`	Scans `/mnt/skills/`, parses SKILL.md frontmatter
`tracker`	Resolves outcome from heuristic + LLM + user signals
`anonymise`	Strips PII via 13 regex patterns, caps at 80 chars
`find-skill`	Merges local + hive data, ranks results
`hive-client`	HTTP client for the 4 hive API endpoints

Privacy by design

Zero PII. No user IDs. No correlation between reports.

No IP addresses stored — Worker does not persist IPs to D1
No session or user identifiers — every report is a standalone row
Summaries are LLM-generated — never raw user input, 13 PII patterns stripped
Tags are author-defined constants — static, safe by construction
No correlation possible — no device fingerprint, no cookie, no auth token

PII patterns detected and redacted

Email addresses Credit card numbers US phone numbers Intl phone numbers IPv4 addresses Social Security numbers macOS user paths Linux user paths Windows user paths UUIDs OpenAI API keys GitHub PATs Slack tokens

Hive API

https://skill-tracker-hive.stas6236.workers.dev

Submit anonymised invocation telemetry.

curl -X POST .../report \
  -H "Content-Type: application/json" \
  -d '{
    "skill_name": "pdf",
    "outcome": "success",
    "tags": ["document-creation"],
    "summary": "Generated a formatted report",
    "timestamp": "2026-03-30T10:00:00Z"
  }'

# Response: {"id": "a1b2c3d4-..."}

Outcomes: success | partial | failure | unknown. Summary max 80 chars.

Search skills by name.

curl -X POST .../query \
  -H "Content-Type: application/json" \
  -d '{"query": "pdf"}'

# {"skills": [{"skill_name":"pdf","total":142,"success_rate":0.87,...}]}

Get aggregated stats for a single skill.

curl .../stats/pdf

# {"skill_name":"pdf","total":142,"successes":124,
#  "failures":12,"partials":6,"success_rate":0.87,
#  "last_seen":"2026-03-30 09:58:56"}

Full skill catalog with success rates.

curl .../catalog

# {"skills": [
#   {"skill_name":"pdf","total":142,"success_rate":0.87,...},
#   {"skill_name":"docx","total":89,"success_rate":0.62,...}
# ]}

Response fields

Field	Type	Description
`skill_name`	string	Skill identifier
`total`	integer	Total invocations
`successes`	integer	Success count
`failures`	integer	Failure count
`partials`	integer	Partial success count
`success_rate`	float	0.0 – 1.0
`last_seen`	string	Last report timestamp

Get started in 10 seconds

claude plugins marketplace add sliday/claude-plugins
claude plugins install alit

Then try:

/find-skill "create PDF"
/find-skill "deploy to Cloudflare"