Junie already reads Agent Skills. Give it one that stops overpaying.

How this maps onto JetBrains AI Assistant

Same rubric. Junie's own Agent Skills support underneath it.

Junie, the agentic coding assistant inside JetBrains AI Assistant, documents support for Agent Skills with three confirmed discovery paths: project-level <projectRoot>/.junie/skills/<name>/, user-level ~/.junie/skills/<name>/ (%USERPROFILE%\.junie\skills\<name>\ on Windows), and the cross-agent .agents/skills/ convention shared with several other tools. Project-level wins on a name collision. That's a direct fit for something already shipped as a SKILL.md — no translation needed.

If no skill applies to a given task, Junie falls back to plain guidelines in .junie/guidelines.md or a root AGENTS.md — but a scored, versioned rubric is exactly what Agent Skills exist for, so the skill install below is the reliable path, not the fallback.

Undercut's tier names — cheap, standard, frontier, apex — are generic on purpose. JetBrains AI Assistant is multi-model, not a single vendor's own models — the plan and IDE decide which providers are on offer. models.md gives the rule of thumb rather than a fixed mapping for Junie specifically: cheap resolves to the smallest/fastest model available in your model picker, standard to your default work model, frontier to the highest-reasoning model on offer, and apex to a single tie-break call on the strongest model available. Nothing here sits between Junie and its model backend — it's a policy file the agent reads at dispatch time, same as any other skill, no proxy, no network hop.

Install for JetBrains AI Assistant

Free. MIT. Nothing installs, nothing proxies your traffic.

Option 1 — project-level skill (recommended)

Junie's own docs confirm project-scoped discovery at .junie/skills/<name>/. Undercut's skill is already named firstpass, so the directory copies straight across — no renaming, no rewriting frontmatter — and it takes priority over a user-level or cross-agent copy if there's ever a name collision.

Option 2 — the cross-agent convention Junie also scans

Junie's docs also list .agents/skills/ as a recognized discovery path — the same convention several other agents in this compatibility bar read. If a repo already carries a skill installed that way for another tool, Junie should pick it up with nothing extra to install.

Validate it yourself

1. 1Read the published results — testing/README.md and the raw JSON in testing/results/
2. 2Reproduce for free, no API key, no spend: node src/main.js --mock
3. 3Open a small task in Junie with the skill installed and check which model tier it picks in the run log — a single-model smoke test is pennies

Zero lock-in. It's a policy Junie reads — remove the skill and you're back to status quo.

Six flags assign a tier. Three triggers are the only way up.

This is a named, versioned spec — not internal logic we could quietly change. When the rubric changes (a new flag, a different cap), the version number moves and the changelog says why.

Every unit of work gets scored against six flags, then assigned the cheapest tier that can plausibly handle it. It only moves up a tier when something objective proves it needs to — a failed check, a measured disagreement, an explicit uncertainty flag. It never moves up on a guess, and it never moves back down mid-task.

 unit of work
      |
  6-flag rubric
      |
  +---+----+
  | 0 flags|--> cheap
  | 1-2    |--> standard
  | 3+/own |--> frontier
  +---+----+
      | fail x2 / disagree / uncertain
      v
  escalate +1 tier
  (residue only)
      |
  still unresolved?
      v
  single batched
  apex tie-break

Why it holds up

Flags steer. Verification and escalation decide.

Even when the rubric flags are scored imperfectly, the escalator still routes work to the right tier almost every time — because a wrong flag only ever costs one extra cheap attempt, never a wrong answer or a big bill.

Honest limits

What this doesn't do.

Not a proxy, not a gateway, doesn't auto-flag in production (Junie scores the flags itself — measured robust to getting them wrong, see above), and doesn't promise a dollar figure for your workload. The Agent Skills discovery paths above come from JetBrains' own docs, researched fresh rather than run end-to-end by us — if you hit a mismatch, see the note in the PR. Full list, with the finance/procurement note and the paid-tier disclosure, is on the main page.