Skip to content
SkillModel-led680 tokens in use

li-inbox

Sorts a pile of LinkedIn messages into lead, recruiter, peer, ask and spam, flags automated sequences and drafts replies for the ones worth sending.

3,760 stars on the repository MIT LicenseJakeschincariol/linkedin-agent-skill · skills/li-inbox
/plugin marketplace add Jakeschincariol/linkedin-agent-skill /plugin install linkedin-agent
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing li-inbox costs

73tokens idle
680tokens in use

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works.

Touches

Nothing detected: no shell, network, file writes, git or credentials in its files.

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Good
  • Frontmatter12/15
  • Structure11/25
  • Examples6/15
  • Output10/15
  • Size15/15
  • Resources10/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Excellent

No findings.

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Needs attention
  • Development activity28
  • Maintenance4
  • Community52
  • Maturity16
  • Momentum40

li-inbox takes the connection requests and DMs you paste in and sorts them into five buckets: lead, recruiter, peer, ask and spam. It prints the counts first, since seeing how much of the inbox is noise is most of the value.

It recognizes the shape of an automated outreach sequence and names the tell that gave it away. Drafts are written only for the buckets that deserve a reply, each one run through the humanizer, and you send them yourself.

What You Get

  • Counts per bucket, such as how many leads, recruiters and spam messages you have
  • A reason for each spam call, naming the tell of the automated sequence
  • Reply drafts for leads, recruiters, peers and asks, with short warm declines when the answer is no
  • A rule for each bucket on how fast to reply and how much to write

Common Use Cases

  • Clearing a backlog of fifty unread DMs in one sitting
  • Deciding whether a recruiter message deserves an answer
  • Declining a request to pick your brain in one warm sentence
  • Spotting a message that is part of a scripted outreach sequence

What Makes It Unique

It sorts before it writes, and writes only where a reply pays off. It also has a description of what automated outreach looks like: a vague invite note, a message minutes after the accept, a calendar link in message one and a bump four days later. It never opens your inbox itself.

Under The Hood

The skill is instructions only, with no scripts. It works from messages and screenshots you paste into the chat and explicitly avoids logging in to your account or reading the inbox with a browser tool. It does not fetch web pages, call APIs or write files, and it passes the drafts through the humanizer skill.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers