# Staff Plus Engineering

> The non-technical and technical-leadership competencies that distinguish Staff, Senior Staff, Principal, and Distinguished engineers from Senior — the things that actually earn promotion past the senior "terminal" level. Use when working on technical leadership, technical strategy/vision, architecture direction, a design doc / RFC / one-pager / Amazon-style narrative, influence without authority, alignment/sponsorship/consensus, leading large ambiguous multi-team efforts, mentorship & sponsorship, the Staff archetypes (Tech Lead, Architect, Solver, Right Hand), glue work, a "staff project" or promotion case, make-vs-buy and tech-debt-as-portfolio, or org-level AI-platform strategy across training/serving/data/infra/cost/safety. Grounded in StaffEng archetypes, Will Larson's "Staff Engineer", and Tanya Reilly's "The Staff Engineer's Path". Ladders/titles vary by company.

- Skill: `sanjeevrg89/staff-plus-engineering` (Agent Skill, multi-file: 5 files)
- Install (CLI): `npx skillmds@latest add sanjeevrg89/staff-plus-engineering`
- Raw SKILL.md: https://api.skillmd.com/api/skills/sanjeevrg89/staff-plus-engineering/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: DevOps & Infra
- Author: sanjeevrg89 (https://skillmd.com/u/sanjeevrg89)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/sanjeevrg89/staff-plus-engineering

---


# Staff+ Engineering

Apply the judgment of a Staff/Principal/Distinguished engineer who has reached org-level impact
through scope, leverage, and influence — not by coding more. The rest of this library is the
*technical* mastery; this skill is how that mastery becomes **world-class impact through others.**
Exact ladders, titles, and level boundaries vary by company — the role's *shape* is universal.

## How to use this skill

1. **Read `staff-plus-engineering-guide.md`** in this directory — the full reference (the
   Senior→Staff discontinuity, the four archetypes, the six core competencies, writing/strategy, how
   to operate, the AI-architect dimension, and anti-patterns). Apply it to the task.
2. For artifacts to imitate — a technical design-doc / RFC template, an engineering-strategy outline,
   and a "staff project" framing tied to AI-platform work — read **`examples.md`**.
3. Adapt to the specific company's ladder, rubric, and doc culture; the competencies are
   industry-general, the mechanics (titles, packet format, who decides) are local — verify them.

## The essentials (full rationale in `staff-plus-engineering-guide.md`)

- **Senior→Staff is a discontinuity, not "more coding."** Senior is judged on *your* output; Staff+
  on what the *org* produces because you were there. Three words: **scope, leverage, impact through
  others.** Your hours are your least important output.
- **Know your archetype** — Tech Lead (a team), Architect (a domain), Solver (the current fire),
  Right Hand (an exec's technical proxy). For AI infra, **Architect** is the default shape.
- **Lead with leverage, not heroics.** The highest-value move is usually a doc that aligns ten teams,
  a strategy that kills a doomed project early, or a mentee who becomes Staff — not your own commits.
- **Pick the right problem.** Staff impact is org-level. The skill is *problem selection* — the
  highest-leverage problem in reach — and the discipline to skip the interesting ones that don't matter.
- **Influence without authority is the whole game.** You have responsibility without command. Currency
  is trust/credibility/social capital: bank it with good calls and real work; spend it on alignment,
  sponsorship, and consensus. Influence by mandate corrodes; being demonstrably right compounds.
- **Writing is the primary tool of leverage.** Design docs / RFCs / one-pagers / Amazon-style
  narratives scale and persist where meetings don't. Write for the audience — execs (BLUF, one page,
  business framing) vs. engineers (depth, trade-offs). If you can't write clearly, your influence is
  capped at the meetings you can attend.
- **Strategy = diagnosis → guiding policy → coherent action** (Rumelt/Larson). Write it yourself;
  pre-socialize with detractors 1:1; sequence the bets; treat it as a living doc. Vision forecasts
  ~2 years out. A strategy that doesn't say *no* to something isn't a strategy.
- **Manage tech debt as a portfolio**, make-vs-buy on TCO + differentiation (not sticker price), and
  build only where it's a genuine edge — most infra is buy/adopt.
- **Multiply others.** Mentorship gives time; **sponsorship** spends *your* capital on someone's
  opportunity — scarcer and worth more. You're a role model whether you want to be or not.
- **Make glue work legible.** The invisible coordinating/unblocking is often the highest-leverage work
  — and gets you passed over if you never build the narrative. Write up what you enabled, in outcomes.
- **Stay technical.** The Staff+ track exists so you *don't* drift to management. Stale judgment makes
  you an expensive manager without a team. Stay close enough to the work that your opinions are earned.
- **Promotions are retroactive.** Staff+ recognizes impact you're already having, evidenced by
  artifacts and testimony — not tenure. The "staff project" manufactures that evidence.
- **AI-architect:** be the one person who reasons across training ↔ serving ↔ data ↔ infra ↔ cost ↔
  safety and makes *defensible* (written-down, alternatives-considered) org-level calls. Favor
  reversible decisions where the ground is moving (it is 2026).

## Related skills

- `[[ml-system-design]]` — the technical design discipline a Staff+ AI-architect drives at org scale.
- `[[aiml-on-kubernetes]]` — the platform substrate; Staff+ owns its direction and make-vs-buy.
- `[[mlops-lifecycle]]` — where platform standards and technical debt accrue across the lifecycle.
- `[[responsible-ai-governance]]` — safety/eval/governance as an org-level technical-health concern.
- `[[go-best-practices]]` — the code-level taste this skill turns into org-level impact.