feat(index): AST-based code indexing and semantic retrieval pipeline

[bug-ops]

Summary

• New zeph-index crate: AST-based code indexing via tree-sitter, semantic retrieval via Qdrant, hybrid query routing, and lightweight repo map generation
• 7 modules (~2,400 LOC): languages, chunker, context, store, indexer, retriever, repo_map
• Feature-gated tree-sitter grammars (Rust, Python, JS/TS, Go, Bash, TOML/JSON/Markdown)
• Dual-write storage (Qdrant vectors + SQLite metadata) with INT8 scalar quantization
• Budget-aware retrieval with query classification (Semantic / Grep / Hybrid)
• Integrated into agent loop via BudgetAllocation.code_context slot and inject_code_context()
• Opt-in via [index] enabled = true config — disabled by default, zero impact on existing functionality
• 54 unit tests, 1190 total workspace tests passing, zero clippy warnings

Architecture

User query
    |
    +-- Contains exact symbol (::, fn, CamelCase, snake_case)?
    |     -> Grep (agent uses bash rg)
    |
    +-- Conceptual query ("how does", "where is")?
          -> Semantic RAG -> Qdrant -> ranked chunks -> inject into context

Dependency graph

A (languages) ---> B (chunker) ---> C (context) ---> E (indexer)
D (storage) --------------------------------> E (indexer)
D (storage) ---> F (retriever)
A (languages) ---> G (repo_map)
E + F + G -----> H (integration)

Validation

Agent	Verdict	Findings
Testing	Pass	54/54 tests, coverage 89-100% on pure-logic modules
Performance	Pass (MVP)	3 post-merge optimizations identified (N+1 queries, JSON round-trip, non_ws_len)
Security	Pass	0 critical/high/medium, 3 low (file size limit, XML escaping, path canonicalization)
Code Review	Approved	Zero blocking issues, 3 important + 5 minor suggestions for post-merge

Test plan

• cargo nextest run -p zeph-index — 54 unit tests
• cargo nextest run --workspace --lib --bins — 1190 tests
• cargo clippy --workspace -- -D warnings — zero warnings
• cargo +nightly fmt --check — clean
• Existing tests unaffected by changes

Closes #179, Closes #180,Closes #181,Closes #182,Closes #183,Closes #184,Closes #185,Closes #186,Closes #187

[codecov-commenter]

Codecov Report

❌ Patch coverage is 70.91575% with 397 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
crates/zeph-index/src/store.rs	38.51%	166 Missing ⚠️
crates/zeph-index/src/indexer.rs	32.53%	85 Missing ⚠️
crates/zeph-index/src/retriever.rs	59.80%	82 Missing ⚠️
crates/zeph-index/src/repo_map.rs	89.94%	17 Missing ⚠️
crates/zeph-core/src/agent.rs	0.00%	16 Missing ⚠️
crates/zeph-index/src/chunker.rs	95.23%	16 Missing ⚠️
crates/zeph-index/src/languages.rs	89.72%	15 Missing ⚠️

@@            Coverage Diff             @@
##             main     #190      +/-   ##
==========================================
- Coverage   86.17%   85.16%   -1.02%     
==========================================
  Files          70       77       +7     
  Lines       19282    20644    +1362     
==========================================
+ Hits        16617    17582     +965     
- Misses       2665     3062     +397     

Files with missing lines	Coverage Δ
crates/zeph-core/src/context.rs	99.47% <100.00%> (+0.01%)	⬆️
crates/zeph-index/src/context.rs	100.00% <100.00%> (ø)
crates/zeph-index/src/languages.rs	89.72% <89.72%> (ø)
crates/zeph-core/src/agent.rs	86.29% <0.00%> (-0.61%)	⬇️
crates/zeph-index/src/chunker.rs	95.23% <95.23%> (ø)
crates/zeph-index/src/repo_map.rs	89.94% <89.94%> (ø)
crates/zeph-index/src/retriever.rs	59.80% <59.80%> (ø)
crates/zeph-index/src/indexer.rs	32.53% <32.53%> (ø)
crates/zeph-index/src/store.rs	38.51% <38.51%> (ø)

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.