Cline already has a native Skills tool that reads SKILL.md. Give it one that stops overpaying.

How this maps onto Cline

A native Skills tool, plus no single house model to map onto.

Cline (formerly Claude Dev) ships a documented Skills feature built around the same SKILL.md layout Undercut ships in: a folder named for the skill, containing a required SKILL.md with YAML frontmatter (name matching the directory, a description that tells Cline when to activate it) plus an optional body of instructions. Cline looks for project-level skills in .cline/skills/ (its own recommended path), and also reads .clinerules/skills/ and .claude/skills/ for compatibility with skills already installed for other tools — plus ~/.cline/skills/ for skills you want available in every project. When a request matches a skill's description, Cline's use_skill tool loads the full SKILL.md body on demand; you can also invoke a skill explicitly as a slash command. That's the exact shape Undercut ships in — no adapter needed. Separately, Cline also reads a workspace .clinerules/ directory of plain instruction files (and auto-detects AGENTS.md, .cursorrules, and .windsurfrules) for general repo conventions that aren't skill-shaped — that mechanism is unrelated to where the Undercut skill itself lives.

Undercut's tier names — cheap, standard, frontier, apex — are generic on purpose, and that matters more for Cline than for a single-vendor tool. Cline isn't built on one company's model family; it's a bring-your-own-key extension that talks to Anthropic, OpenAI, Google, local and open-weight models, and others, all through the same interface. models.md gives a rule of thumb rather than one fixed mapping for exactly that reason: cheap resolves to the smallest/fastest model you've configured in Cline, standard to your default work model, frontier to the highest-reasoning model on your roster, and apex to a single batched tie-break call on whichever model you'd trust most with the hardest 1% of the work.

Nothing about this sits between Cline and whichever model backend you've configured. It's a policy file Cline's Skills tool reads at dispatch time — no proxy, no network hop, nothing to route around if you remove the skill directory.

Install for Cline

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

Option 1 — skills CLI

This is the same generic skills CLI referenced on the main site. We haven't independently verified which of Cline's recognized skill directories it targets — if it writes to .claude/skills/, that's one of Cline's documented compatibility paths and needs no change; if you're unsure, use Option 2 below and place it under Cline's own recommended .cline/skills/ directly.

Option 2 — copy the skill directory directly

Clone or download undercutsh/firstpass, then copy skills/firstpass/ into .cline/skills/firstpass/ at your project root (or ~/.cline/skills/firstpass/ to make it available in every project). Cline also reads .claude/skills/ and .clinerules/skills/, so it works dropped in either of those instead.

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. 3Run live with whichever provider you've already configured Cline against — a single-vendor smoke test is pennies

Zero lock-in. It's a policy Cline's Skills tool reads — delete the directory 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 (Cline scores the flags itself — measured robust to getting them wrong, see above), and doesn't promise a dollar figure for your workload. It also doesn't pick a model for you — on a bring-your-own-key host like Cline, resolving each tier to a concrete model is still a choice you make in Cline's own model settings. The exact skill-directory precedence and use_skill activation behavior described above come from Cline's published docs; we haven't independently re-tested every path and edge case end-to-end ourselves. Full list of general limits, with the finance/procurement note and the paid-tier disclosure, is on the main page.