compiler-orchestrator

An end-to-end workflow for advancing the React Compiler’s Rust rewrite. It checks the current porting progress, fixes broken stages, moves additional stages from TypeScript, reviews the code, and commits the work in a loop.

In plain words
What is it for?
Use it to continue the Rust compiler port, inspect its current status, start from a chosen compiler stage, fix failures, and review or commit progress.
Why use it?
A compiler rewrite has many dependent stages, so it is easy to lose track of what is ready, what is failing, and what should come next. This workflow coordinates those activities around the compiler’s defined pass order.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/react/react/compiler-orchestrator
Any agent
npx skills add react/react --skill compiler-orchestrator
Clone the repo
git clone --depth 1 https://github.com/react/react

Made for: Claude Code, Codex.

Per session 38 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,382 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00038 $0.03382
Opus 5 $0.00019 $0.01691
Sonnet 5 $0.00008 $0.00676
Haiku 4.5 $0.00004 $0.00338

Measured yesterday against content hash 8f4e69d1d932, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

compiler-orchestrator scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

compiler/.claude/skills/compiler-orchestrator/SKILL.md · 274 lines

How it starts

The opening of the file, as written. The whole thing — 274 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Compiler Orchestrator

Automatically drive the Rust compiler port forward by discovering the current state, fixing failures, porting new passes, reviewing, and committing — in a continuous loop.

Arguments:

  • $ARGUMENTS: Optional. A pass name to start from, or status to just report current state without acting.

Pass Order Reference

These are the passes in Pipeline.ts order, with their exact log names:

# Log Name Kind Notes
1 HIR hir
2 PruneMaybeThrows hir Validation: validateContextVariableLValues, validateUseMemo after
3 DropManualMemoization hir Conditional
4 InlineImmediatelyInvokedFunctionExpressions hir
5 MergeConsecutiveBlocks hir
6 SSA hir
7 EliminateRedundantPhi hir
8 ConstantPropagation hir
9 InferTypes hir Validation: validateHooksUsage, validateNoCapitalizedCalls after (conditional)
10 OptimizePropsMethodCalls hir
11 AnalyseFunctions hir
12 InferMutationAliasingEffects hir
13 OptimizeForSSR hir Conditional: outputMode === 'ssr'
14 DeadCodeElimination hir
15 PruneMaybeThrows (2nd) hir Reuses existing fn, just needs 2nd call + log in pipeline.rs
16 InferMutationAliasingRanges hir Validation block (8 validators) after (conditional)
17 InferReactivePlaces hir Validation: validateExhaustiveDependencies after (conditional)
18 RewriteInstructionKindsBasedOnReassignment hir Validation: validateStaticComponents after (conditional)
19 InferReactiveScopeVariables hir Conditional: enableMemoization
20 MemoizeFbtAndMacroOperandsInSameScope hir
-- outlineJSX hir Between #20 and #21, conditional: enableJsxOutlining, no log entry
21 NameAnonymousFunctions hir Conditional
22 OutlineFunctions hir Conditional
23 AlignMethodCallScopes hir
24 AlignObjectMethodScopes hir
25 PruneUnusedLabelsHIR hir
26 AlignReactiveScopesToBlockScopesHIR hir
27 MergeOverlappingReactiveScopesHIR hir
28 BuildReactiveScopeTerminalsHIR hir
29 FlattenReactiveLoopsHIR hir
30 FlattenScopesWithHooksOrUseHIR hir
31 PropagateScopeDependenciesHIR hir
32 BuildReactiveFunction reactive
33 AssertWellFormedBreakTargets debug Validation
34 PruneUnusedLabels reactive
35 AssertScopeInstructionsWithinScopes debug Validation
36 PruneNonEscapingScopes reactive
37 PruneNonReactiveDependencies reactive
38 PruneUnusedScopes reactive
39 MergeReactiveScopesThatInvalidateTogether reactive
40 PruneAlwaysInvalidatingScopes reactive
41 PropagateEarlyReturns reactive
42 PruneUnusedLValues reactive
43 PromoteUsedTemporaries reactive
44 ExtractScopeDeclarationsFromDestructuring reactive
45 StabilizeBlockIds reactive
46 RenameVariables reactive
47 PruneHoistedContexts reactive
48 ValidatePreservedManualMemoization debug Conditional
49 Codegen ast

Read the full file on GitHub · 274 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 274 lines · 38 tokens per session scan A 8f4e69d1d932

Subscribe to this mod's changes

compiler-orchestrator is a skill published in the GitHub repository react/react (248,328 stars, last pushed 4d ago), licensed MIT. It adds 38 tokens to every session and 3,382 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.