frontier-infra/proctor/goal-contract

Goal Contract

Turn a rough task into a locked contract before any code is written. Adds the missing third leg — cost — to the usual goal/acceptance/non-goals, so work can neither drift off-objective nor balloon past its worth.

Why — the three legs

Every unit of work needs all three. Most setups have the first two and miss the third:

• Context — the situation (claude-mem, vault, ops already provide this).
• Direction — Goal + Acceptance + Non-goals.
• Cost — Budget + Tripwire. This is the leg usually missing. An agent has no

intrinsic price on its own effort, so without a cost leg it both wanders (drift) and over-builds (a 1-hour job becomes a day). Same hole, two symptoms.

Workflow

1. Take the rough statement of the next task/phase. Ask at most ONE clarifying

question, and only if the goal is genuinely ambiguous. Do not interrogate.

1. Draft the contract using the template below. Acceptance must be observable;

non-goals concrete; budget real (not "as needed").

1. Offer a distilled version — same contract, tighter. Let the user pick full or distilled.
2. Emit the paste-ready LOCKED CONTRACT block. The user pastes it back to start work.

Contract template

• GOAL — one sentence, an outcome (not a method).
• ACCEPTANCE — 2–4 bullets, each observable/testable ("endpoint returns 200", "test passes", "phase < 200ms").
• NON-GOALS — concrete things NOT to touch this round (named files, refactors, features, "nice-to-haves").
• BUDGET — an effort ceiling. Pick what fits: ~time (e.g. ~1h), ≤N files, or "smallest change that meets acceptance".
• TRIPWIRE — if budget is hit, OR scope wants to expand, OR a refactor outside the goal tempts → STOP and report; don't grind.

Paste-ready output

=== LOCKED CONTRACT ===
GOAL: ...
ACCEPTANCE:
  - ...
NON-GOALS:
  - ...
BUDGET: ...
TRIPWIRE: At the budget — or any scope/non-goal temptation — STOP and report before
continuing. No grinding, no gold-plating.
RULES: Restate this contract before the first action. Check every step against GOAL +
NON-GOALS. Report against ACCEPTANCE at the budget and before declaring done.
=== Work only to this contract. ===

For the agent executing a locked contract

• Restate GOAL + NON-GOALS in one line before the first action.
• The smallest change that meets ACCEPTANCE wins. Extra tests, abstractions, or

cleanup not named in ACCEPTANCE are non-goals — adding them is overcorrection.

• Pre-existing brokenness is the one carve-out from "smallest change." If you walk

into a failing test, type debt, dead code, or a stale doc — in a file you're touching, or otherwise cheap — fix it; don't note-and-leave. If the honest fix would blow the budget/scope, that's a tripwire: STOP, surface it with the measured cost, and let the user choose (expand now vs. explicitly defer). This is distinct from gold-plating above: adding things nobody asked for is still overcorrection; leaving existing brokenness noted-but-unowned is not allowed. "The linter will catch it" is banned. Owned means you are taking the problem into this contract to fix it before done — not that you appended a Follow-Up and walked away. If something truly cannot be fixed inside this contract, call it deferred/operator-blocked, with owner + next action + proof, and state whether the goal is blocked.

• Hit the budget, or tempted past scope? Stop and report — never unilaterally decide to go bigger.

Example

Rough input: "make the app faster, the startup is slow"

=== LOCKED CONTRACT ===
GOAL: Cut cold-start latency so the service is ready to serve in under a second.
ACCEPTANCE:
  - startup < 1s (startup benchmark)
  - endpoints unchanged (no behavior change)
NON-GOALS:
  - the request cache, the installer, the dashboard, the migration runner
BUDGET: ~half-day, ≤4 files, behind the existing loader.
TRIPWIRE: if it needs a new subsystem or touches >4 files, STOP and report. No grinding.
RULES: Restate before starting. Check each step against GOAL + NON-GOALS. Report at budget and before done.
=== Work only to this contract. ===