Crate ainl_context_compiler 

AINL Context Compiler — LLM context-window assembly

Phase 6 of SELF_LEARNING_INTEGRATION_MAP.md.

Multi-segment, role-aware, question-aware prompt orchestration with progressive-enhancement capability tiers (heuristic → LLM-driven anchored summarization → embedding-based relevance).

Boundary vs ainl-context-freshness

These two crates share a name prefix but solve different problems:

Crate	Lifecycle phase	“Context” means
ainl-context-freshness	Pre-tool execution policy gate	the agent’s knowledge of the world (repo/index state vs HEAD)
ainl-context-compiler (this crate)	Prompt assembly / window management	the LLM’s input context window (prompt bytes about to be sent)

ainl-context-compiler consumes ainl-context-freshness as a per-segment rank-down signal (stale segments are ranked lower) — see relevance::HeuristicScorer.

See docs/ainl-crates-overview.md for the broader ainl-* family map.

Design tiers (auto-detected at runtime)

• Tier 0 — Heuristic (always available): question-token overlap × recency × freshness; per-segment compression via ainl_compression.
• Tier 1 — Anchored summarization (M2): when a summarizer::Summarizer is injected, older history collapses into a structured AnchoredSummary (Factory.ai pattern).
• Tier 2 — Embedding rerank (M3): when an embedder::Embedder is injected, segments are reranked by cosine similarity to the latest user message.

Each tier auto-degrades on per-call failure; the system never blocks on optional capabilities.

Telemetry sink (canonical pattern from §15.4)

Mirrors ainl_compression::CompressionTelemetrySink exactly. Hosts implement ContextEmissionSink once and pass it via orchestrator::ContextCompiler::with_sink; everything downstream just emits structured events.

Telemetry field names come from ainl_contracts::telemetry constants prefixed CONTEXT_COMPILER_* so dashboards, Prometheus exporters, and CI gates reference them consistently across hosts.

Re-exports

pub use budget::BudgetPolicy;

pub use capability::CapabilityProbe;

pub use capability::Tier;

pub use embedder::cosine;

pub use embedder::Embedder;

pub use embedder::EmbedderError;

pub use embedder::PlaceholderEmbedder;

pub use failure_recall::memory_block_for_user_query;

pub use mcp_ainl_prompt::mcp_ainl_run_adapters_cheatsheet_segment;

pub use mcp_ainl_prompt::MCP_AINL_RUN_ADAPTERS_CHEATSHEET;

pub use metrics::ContextCompilerMetrics;

pub use metrics::SegmentMetrics;

pub use orchestrator::ComposedPrompt;

pub use orchestrator::ContextCompiler;

pub use relevance::HeuristicScorer;

pub use relevance::RelevanceScore;

pub use relevance::RelevanceScorer;

pub use segment::Role;

pub use segment::Segment;

pub use segment::SegmentKind;

pub use summarizer::AnchoredSummary;

pub use summarizer::AnchoredSummarySection;

pub use summarizer::Summarizer;

pub use summarizer::SummarizerError;

pub use trajectory_recap::format_trajectory_recap_lines;

Modules

budget

Token budget policy.

capability

Capability tier auto-detection.

embedder

Tier 2 embedding surface (M3).

failure_recall

Optional failure recall → memory-block segment (sources-failure-warnings feature).

mcp_ainl_prompt

Optional compressed host hint for mcp_ainl_ainl_run adapter registration.

metrics

Per-call telemetry struct emitted alongside crate::ComposedPrompt.

orchestrator

ContextCompiler::compose — orchestrates segment selection, budget allocation, compaction, and per-segment compression, returning a ComposedPrompt plus telemetry.

relevance

Relevance scoring (Tier 0 heuristic; Tier 2 embedding rerank in embedder.rs).

segment

Prompt-window segments.

summarizer

Summarizer trait + structured anchored-summary types.

trajectory_recap

Optional trajectory recap lines for prompt injection (sources-trajectory-recap feature).

Enums

ContextCompilerEvent

Structured event emitted by ContextCompiler during prompt composition.

Traits

ContextEmissionSink

Optional structured telemetry sink for context-compiler events.

Type Aliases

SinkRef

Convenience type alias used throughout the crate.