Slice: Author and prove Jim's first Edict operation ReplaceRange

[flyingrobots]

Slice

Author and prove ReplaceRange.edict as Jim's first real application operation.
This proves an operation law and compiler/runtime substrate; it does not by
itself prove the final Jim.edict observer composition.

Parent: #295.

Artifact boundary

• jedit.text.schema@1 is the application ABI and fact/codec/identity contract.
• jedit.text.ReplaceRange.oracle@1 is independent expected-behavior evidence.
• ReplaceRange.edict is the authored executable semantics.
• Edict Core/Target IR plus an independently verified package is executable
  meaning.
• Echo receipts are per-execution evidence.

Neither the schema nor oracle is a program. Echo must not synthesize a package
or evaluator from them.

Required first RED

Check in the exact Jim-owned source and complete lawpack closure, then build it
through Edict's public application-build boundary. The first honest failure
routes work to its owner:

• missing expressivity or Core representation -> Edict;
• source that cannot lower -> Edict target adapter/lawpack;
• verified generic profile unsupported at runtime -> Echo;
• execution disagrees with the oracle -> triangulate source/compiler/runtime.

Do not start by reverse-engineering an Echo evaluator from the schema/oracle.

Acceptance criteria

• ReplaceRange.edict and its complete closure are checked in under a
  Jim-owned application root.
• Edict's public build emits exact Core, Target IR, package, manifest,
  codecs, and structurally separate verification evidence.
• Echo consumes the exact compiler-produced package without reconstruction.
• Echo production code uses only generic bounded-program primitives and
  contains no ReplaceRange, rope, split/join/balance, Buffer, or
  TextWindow branch or intrinsic.
• One admitted execution produces the corpus-defined persistent-rope
  consequence or typed obstruction with basis, budget, footprint, Tick, WAL,
  receipt, reading, and recovery evidence.
• All retained oracle cases agree through a separately implemented runtime
  path that shares no native Jedit planner or algorithm.
• Negative gates reject handwritten packages, caller-authored graph patches,
  MutationPlan, native callbacks, fake transports, and compatibility fallback.
• Any direct operation client/runner is explicitly test-only conformance
  scaffolding and is unreachable from final production composition.

Source location

Use a Jim-owned application root such as:

edict/replace-range/edict.application.json
edict/replace-range/src/ReplaceRange.edict
edict/replace-range/vendor/jedit-text/...

The oracle remains under contracts/jedit/lawpacks/replace-range-v1/ and is not
part of executable semantic input.

Dependencies

• Freeze ReplaceRange canonical fact law and oracle #293 published the frozen schema and oracle.
• Compile bounded control flow and digest-bound pure lawpack helpers into Core edict#192 owns bounded control-flow and digest-bound pure-helper
  lowering into Core. Current public-build probes stop there before provider
  invocation.
• Interpret compiler-produced bounded Edict graph programs echo#684 owns only generic bounded-program interpretation proven
  necessary by the compiler-produced package.

Non-goals

• No Jim.edict observer cutover in this slice.
• No buffer creation, checkpoint, save/export, causal inverse, multi-range edit,
  rope intrinsic, unrestricted VM, child lane, or wormhole.

[coderabbitai]

🔗 Related PRs

#9 - test(stack): add Stack Witness 0001 jedit transport witnesses [closed]
#27 - feat(echo): wire trusted runtime lifecycle into jedit [closed]
#268 - WF-0154 - E-Brake: Observed Absurdity Fixes (cycle) [merged]
#280 - Add runtime-backed causal range evidence [merged]
#293 - Freeze ReplaceRange canonical fact law and oracle [open]

---

📝 Issue Planner

_{Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.}

• Create Plan

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!