/expand

/expand is your profile's archaeologist. You've done more than your profile says — a course here, a side project there, a tool you picked up on the job. /expand goes looking for those competencies across everything it can reach and adds them to your profile. It never edits or deletes what's already there, so running it is always safe.

Where it looks

1. Your documents — documents/cv, linkedin, diplomas, references.
2. Your GitHub — every public repository, pinned or not: it reads the READMEs and notes the languages and topics.
3. Your other links — portfolio, Kaggle, Google Scholar, a personal site.

How it turns a find into a skill

For each thing it discovers, it searches the web two ways at once:

• Direct lookup — the actual syllabus of a named course, the skills behind a certification, the docs for a tool.
• Inference — what methods and toolchains that kind of work normally requires.

Inferred items are always labelled as inferred, so you can tell a documented skill from a reasoned one.

You see everything before anything is written

you:    /expand
claude: Read your profile. Scanning documents/, GitHub (14 repos), and your Kaggle…

        Competency Map (new items only)
        ── Technical — Primary ────────────────────────────────────
        PyTorch          GitHub: vision-experiments        direct
        ── Methods & Practices ────────────────────────────────────
        MLflow tracking  Coursera: MLOps Specialisation    both
        ── Domain Knowledge ───────────────────────────────────────
        Time-series forecasting   GitHub: demand-forecaster inference

        Add these? (all / review each / skip / skip a group)
you:    all
claude: Added 7 competencies across 2 files. Report below.

Your options are all, review (walk group by group), skip, or skip just one category. Nothing is written until you say so.

Additive and idempotent — by design

• It only adds. Existing lines are never changed or removed.
• Every addition carries a source note, like (Coursera — Deep Learning Specialisation) or (GitHub — repo: demand-forecaster). That note is also how /expand stays idempotent: run it again next month and it brings in only what's new.
• Nothing is invented. Every competency traces to a real source or is flagged as an inference, and any document it can't read is reported rather than guessed at.

What you get at the end

An expansion report: what was added to each file (with sources), every source it scanned and how many competencies each yielded, anything it skipped and why, and any items worth a manual look. A good next move afterwards is /upskill — re-check your gaps now that your skill set is richer.

/expand can also be launched from the dashboard Console; it runs this same command.