umbrova/ladderline

A local-first CLI for logging promotion evidence against your team's career ladder — plain files, no server, no AI-invented claims. - umbrova/ladderline

0 stars · 0 forks · TypeScript Topics: cli, typescript, promotion, engineering-management, local-first, career-ladder

README

Ladderline

Track career-ladder evidence, locally, as it happens.

There's also a local dashboard — see the full tour with screenshots → Every engineering manager knows the feeling: promo season arrives, and you're trying to reconstruct six months of a report's growth from memory, old Slack messages, and a vague sense that something good happened in March. Ladderline fixes that by giving you a place to jot a two-line note the moment something evidence-worthy happens — tagged against your team's actual career ladder — so a real, dated record builds itself quietly all year. When it's time to write the case, you're assembling, not remembering.

Local-first. Everything lives in plain YAML/Markdown files on your machine — no server, no account, no database, nothing to trust with sensitive people-data but yourself. Never invents. Cases are assembled only from what you actually logged. Gaps are shown honestly, not smoothed over. Plain files, no lock-in. Notes are Markdown with YAML frontmatter — open them in any text editor, or Obsidian, or grep them. Nothing about this tool traps your data.

Install npm install -g @umbrova/ladderline Or try it without installing anything: npx @umbrova/ladderline init Quick start

Create a workspace in the current folder

Define a review period

Start tracking someone against the bundled default ladder

Log evidence the moment something happens — takes 10 seconds

See it all in the local dashboard

When review time comes, assemble a case from everything you logged

Tracking direct reports heading toward a promotion Your own case, if you're building evidence for your own next level Peer or 360 contributions — logging what you noticed about someone outside your reporting line, for when their manager asks

Track anyone with --as report | self | mentee | cross-team | peer — the mechanism is the same regardless of the relationship. Commands

Command What it does

ladderline init [--demo] Create a new workspace, optionally seeded with a demo person and one deliberate evidence gap

ladderline ladder list Show ladders registered in this workspace

ladderline ladder add <file> Register a ladder file

ladderline ladder remove <file> [--force] Remove a registered ladder

ladderline track <name> --ladder <file> --as <relationship> Start tracking a person

ladderline untrack <name> [--purge] Stop tracking (archives by default; --purge permanently deletes, asks you to type the name to confirm)

ladderline note <text> --person <name> [--tag <id> | --notag] [--date <date>] Log a single evidence note

ladderline note-delete --person <name> --date <date> [--tag <id> | --notag] [--filename <name>] Delete a single note (always asks to confirm)

ladderline cycle add <name> --start <date> --end <date> Define a review period

ladderline cycle list Show defined cycles

ladderline cycle remove <name> Remove a defined cycle

ladderline case [name] --cycle <name> [--format docx|md] [--all] [--as <relationship>] [--prompt] Assemble a case — one person, or --all for everyone

ladderline notag list [--person <name>] List notes not yet mapped to a competency

ladderline export [--person <name>] [--cycle <name>] [--since <date>] Zip the workspace, or a filtered slice of it

ladderline import <file.zip> [--force] Restore or merge a workspace from an export

ladderline dashboard [--port <port>] Launch the local web dashboard

Every command supports --help. Full details, file formats, and conventions live in the wiki. The dashboard ladderline dashboard starts a local web server (default http://localhost:4200, reachable only on your machine) with five views:

Person — one person at a time, each competency as a row, with note counts and how fresh the evidence is Team grid — everyone at a glance, competencies as columns, for spotting gaps before calibration Notes — every logged note, filterable and expandable to its raw file Insights — coverage percentage, what's going stale, your own logging cadence, and cycle readiness Docs — the same reference material as the wiki, rendered locally so it works fully offline

A small badge in the top bar surfaces when something needs attention — the same signal also shows up as a one-line nudge before any command's output, right in your terminal. Why plain files Every note is Markdown with YAML frontmatter: --- person: john-doe tag: technical-direction date: 2026-02-10 cycle: 2026-Q1 --- Pushed back on the caching design in the payments migration doc, proposed the write-through approach, and 3 other teams adopted it after her RFC review. You can write these by hand — in Obsidian, in any editor, however you like. Ladderline doesn't care how a note was created, only that it's a valid file in the right place. Full conventions are documented in the wiki. Feedback

Bugs or ideas → Issues or Discussions See CONTRIBUTING.md for how the codebase is organized

No telemetry, ever — nothing about your usage leaves your machine. License MIT — see LICENSE.

source