Qoder already reads SKILL.md files natively. Give it one that stops overpaying.

How this maps onto Qoder

.qoder/skills/ is Qoder's own native SKILL.md path.

Per Qoder's own Skills documentation, which the docs say applies identically to Qoder IDE and Qoder CLI, a skill is a folder containing a SKILL.md with required YAML frontmatter — name (lowercase letters, numbers, and hyphens, max 64 characters) and description (max 1,024 characters, used by the model to decide when to load the skill). Skills live at .qoder/skills/<name>/SKILL.md for a project or ~/.qoder/skills/<name>/SKILL.md globally, with one documented precedence quirk worth flagging: unlike most other agents on this site, Qoder's user-level skill wins over a project-level skill of the same name, not the other way around. The full body loads only when Qoder decides the current task matches the description; you can also invoke a skill manually with /skill-name, and an existing session needs /skills reload to pick up a skill added mid-session.

Qoder also documents a separate, always-on Rules mechanism — markdown files under .qoder/rules/, shared via git, capped at 100,000 total characters across all active rule files (excess is truncated), with four apply modes: Always Apply, Model Decision, Specific Files (glob patterns), or Apply Manually via @rule in chat. Separately, Qoder's docs say Qoder now recognizes a root AGENTS.md automatically with no extra configuration — but rules content takes precedence over AGENTS.md content if the two conflict. A scored routing rubric is conditional, invoked-by-relevance guidance rather than a blanket always-on preference, so the native Skills path — not Rules or AGENTS.md — is the documented fit this page installs to.

Undercut's tier names — cheap, standard, frontier, apex — are generic on purpose. Qoder isn't tied to one model vendor, so models.md resolves each tier to a real model for whichever family you've configured. Nothing about this sits between Qoder and your model provider — it's a policy file the agent reads at dispatch time, no proxy, no network hop, nothing to route around if you remove the skill folder.

Install for Qoder

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

Option 1 — copy the skill directory directly (recommended, confirmed path)

Clone or download undercutsh/firstpass, then copy skills/firstpass/ into .qoder/skills/firstpass/ at the project root. The frontmatter already matches what Qoder's Skills docs require (a name matching the folder, a description under the 1,024-character cap), so nothing needs rewriting.

Option 2 — install once, for every project

Qoder also reads a user-level skills root: ~/.qoder/skills/<name>/SKILL.md. Per Qoder's own docs, a user-level skill takes priority over a project-level skill of the same name — the reverse of the usual "workspace wins" rule — so if you also install a project copy later, the global one is what actually loads.

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 Qoder on a repo with the skill installed and confirm it references the routing policy before starting mechanical work

Zero lock-in. It's a policy Qoder's agent reads — delete the folder 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 enforce anything at the network layer, and doesn't auto-flag in production — Qoder's agent scores the flags itself (measured robust to getting them wrong, see above). Doesn't promise a dollar figure for your workload — savings are workload-dependent, and real workloads escalate more than benchmarks. Every figure on this page is an observation from our tasks and our graders, not a guarantee about your codebase — verification is not proof, it's a reproducible data point you can run yourself. The Skills, Rules, and AGENTS.md mechanics described above come straight from Qoder's own docs, including the user-overrides-project precedence quirk, but we haven't independently run this skill inside a live Qoder session end-to-end — treat that specific claim as unconfirmed until you check it yourself with the steps above. Full list of general limits, with the finance/procurement note and the paid-tier disclosure, is on the main page.