network-jobs-import
Skill0 starsChanged 3 months agoMITView on GitHub
Import a LinkedIn connections ZIP into the local Network Jobs graph. Use when the user provides a LinkedIn data export ZIP or Connections.csv and wants companies/people available for careers discovery.
In this repository7 other files it gives its agents.
- AGENTS.mdAGENTS.md
- worker/AGENTS.mdAGENTS.md
- careers-discoverSkill
- intro-email-generatorSkill
- jobs-ingestSkill
- network-jobs-setupSkill
- network-jobsSkill
The file
We split this skill into the rules it gives an agent
8 rules. About 0 are common, 8 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈8
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 27
One question per call: Ask it, wait for the answer, then ask the next.
- Distinctive: line 28
Never bundle two questions into one prompt or one tool call, even when the tool accepts several.
- Distinctive: line 33
Put the deciding data in the option label (connection counts, both names), never in surrounding prose the user has to correlate.
- Distinctive: line 35
Always keep a free-text / "none of these" path.
- Distinctive: line 36
No such tool in this host?
- Distinctive: line 81
Do not upload the ZIP or CSV anywhere.
- Distinctive: line 82
Do not invent companies or people not in the file.
- Distinctive: line 83
Prefer the helper over ad-hoc Python/Node one-offs unless the helper fails — then fix/extend the helper.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Network Jobs Import
Turn a LinkedIn connections export into local connections.json + companies.json.
Read SCHEMA.md first.
Inputs
- Path to LinkedIn ZIP (preferred) or a bare
Connections.csv - Optional:
--min-count N(default 1) — drop companies with fewer than N connections when summarizing
Asking the user
Whenever the next step is picking from a known set, use the host's choice tool —
AskUserQuestion (Claude Code), question (opencode), ask_user_choice, or the
harness equivalent — instead of a prose list.
- One question per call. Ask it, wait for the answer, then ask the next. Never bundle two questions into one prompt or one tool call, even when the tool accepts several.
- Recommend, and say why. Lead with the option you would pick and a one-line reason ("Recommended: merge Meta → Facebook — 38 connections otherwise split across two rows").
- Put the deciding data in the option label (connection counts, both names), never in surrounding prose the user has to correlate.
- Always keep a free-text / "none of these" path.
- No such tool in this host? Use a short numbered list, recommendation first, never more than 4 options.
Workflow
- Confirm data home —
NETWORK_JOBS_HOMEor~/.network-jobs/. - Run the parser (deterministic — do not reimplement CSV parsing in prose):
DATA="${NETWORK_JOBS_HOME:-$HOME/.network-jobs}" # always resolve from the env; never export a literal DATA
# Prefer the CLI when available:
network-jobs import "/path/to/linkedin.zip"
# Or call the helper that ships with this skill (works when the skill is symlinked):
HELPER="$(dirname "$0")/helpers/parse-linkedin.sh"
# When running from an agent without dirname context, resolve via suite root:
SUITE="$(cat "${NETWORK_JOBS_HOME:-$HOME/.network-jobs}/suite-root" 2>/dev/null || true)"
SUITE="${NETWORK_JOBS_SUITE:-${SUITE:-}}"
HELPER="${HELPER:-$SUITE/skills/network-jobs-import/helpers/parse-linkedin.sh}"
"$HELPER" "/path/to/linkedin.zip" --out "$DATA"- Report
- Total connections kept
- Companies after ignore-list filtering
- Top 15 companies by connection count (name + count + sample people)
- Optional fuzzy merge — If the user asks, or obvious duplicates appear (e.g. “Meta” / “Facebook”), propose the merges as one multi-select — each option labelled
Meta → Facebook (38 connections)with your recommended merges marked — write the accepted ones to$DATA/config/company-overrides.jsonas{ "linkedinName": "...", "matchTo": "..." }, then re-run the helper. - Hand off — Suggest
careers-discoveron top companies by count.
Helper behavior
helpers/parse-linkedin.sh:
- Extracts
Connections.csvfrom ZIP (or reads CSV directly) - Skips LinkedIn note preamble; finds header with First Name + Company
- Applies ignore lists from
$DATA/config/(seeded fromassets/on setup) - Applies company-overrides
- Writes:
$DATA/connections/connections.json$DATA/companies/companies.json$DATA/logs/import-<timestamp>.jsonsummary
Non-negotiables
- Do not upload the ZIP or CSV anywhere.
- Do not invent companies or people not in the file.
- Prefer the helper over ad-hoc Python/Node one-offs unless the helper fails — then fix/extend the helper.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

