v0.7.0

Changelog

• 1b30298 Add filesystem fallback to codeindex_read for missing files

v0.6.0

v0.6.0 — Exact grep semantics, token-efficient output, hierarchical ignore files

This release replaces the search engine and fixes two silent correctness bugs that could mislead AI agents, while cutting both memory usage and output token cost.

🔍 Search engine rewritten: exact grep semantics

The Bleve token-based engine had silent false negatives:

• substring queries inside identifiers returned nothing (Mutex missed sync.RWMutex)
• /regex/ queries matched files but returned zero lines (the line extractor treated the pattern as a literal)
• multi-word queries OR-ed tokens while the line matcher required the full string

Search is now a direct in-memory line scan compiled to a single RE2 matcher — results are grep-identical with full recall, and memory drops to roughly the corpus size (the Bleve dependency tree is gone).

Query formats:

Format	Behavior
plain text	literal substring, case-insensitive
"quoted"	exact literal, case-sensitive
/regex/	RE2 regular expression per line, case-insensitive by default

🪙 Token-efficient output

• Overlapping match contexts merge into hunks (N: = match, N- = context, -- separator) — no more duplicated context lines
• maxMatchesPerFile display cap (default 10) with ... +N more matches
• New outputMode: content (default), files (paths only), count (paths + match counts) — use files/count when locations suffice
• contextLines: 0 now works (was silently reset to 2)
• New caseSensitive option
• Explicit notice when the file limit truncates results

📂 Hierarchical ignore files

Nested .gitignore / .claudeignore files at every directory level are now respected with git's precedence rule (closest file wins, negations included). Previously only the root files were loaded — monorepo subprojects leaked build output into the index.

.env.* and *.env are now ignored by default so secrets never enter the index (--force-include overrides intentionally).

🔄 Index consistency hardening

• Newly created directory trees (copy, unzip, git checkout) are watched recursively and files already inside are indexed — closes a race where content written before the watch attached was never indexed
• Periodic sync verification is on by default (every 300s, --sync-interval 0 disables) as a self-healing net
• .gitignore/.claudeignore changes now re-sync the index (newly-ignored files dropped, newly-unignored indexed)

🛠 MCP improvements

• All read-only tools declare readOnlyHint so clients can skip permission prompts
• codeindex_status reports version, watched directory count, and sync interval
• Tool descriptions document the new grep semantics and token-saving guidance
• Logging defaults to off (no log file unless --log-enabled)

⬆️ Upgrading

Drop-in replacement — rebuild or download the new binary and restart your MCP client session. Tool parameters are backward compatible; the search output format changed (hunks with N:/N- markers).

v0.5.1

Changelog

• f979cd5 Add offset/limit parameters to codeindex_read for partial file reading
• 6db85f3 Bump version to 0.5.1

v0.5.0

Changelog

• cd4b6f0 Minimize tool output for AI token efficiency, bump to 0.5.0

v0.4.0

Changelog

• cf20300 Add GoReleaser config and release workflow for cross-platform binaries
• de9e427 Add badges to README (CI, Go Report Card, pkg.go.dev, license, MCP, Claude Code)
• 552a056 Add register subcommand for auto-registration in Claude Code, bump to 0.4.0
• 3888b79 Fix flaky Duration assertion in empty directory sync test
• ce91479 Simplify Quick start section in README

v0.3.0

What's New

Toggleable logging

• New --log-enabled flag (default: true). Set to false to disable all log output — no log file is created, no stderr output.
• Fixed log file handle leak — the log file is now properly closed on exit.

Periodic index sync verification

• New --sync-interval flag (in seconds, default: 0 = disabled).
• When enabled, a background goroutine periodically compares the filesystem with the index and fixes discrepancies:
  • Missing files — on disk but not indexed → re-indexed
  • Stale files — in index but deleted from disk → removed
  • Modified files — ModTime changed → re-indexed
• All discrepancies are logged in detail (path, type, action taken).

Other

• Input validation: --sync-interval rejects negative values.
• 9 new unit tests for sync verification (happy paths + edge cases).
• Documentation updated with new flags, examples, and project structure.

Full Changelog

v0.2.0...v0.3.0

v0.2.0

What's New

--force-include CLI flag

Override all exclude rules (default patterns, .gitignore, .claudeignore, --exclude) for specific file patterns.

# Index log files even though *.log is excluded by default
./codeindex-mcp --root /project --force-include "*.log"

# Multiple patterns (additive)
./codeindex-mcp --root /project --force-include "*.log" --force-include "vendor/*.go"

Key behaviors:

• Force-include takes highest priority over all exclude rules
• Multiple --force-include flags are additive
• Binary detection and file size limits still apply (safety)
• .git directory is always skipped regardless of force-include
• Directory pruning is suppressed for paths that could contain force-included files

Other

• Full test coverage for force-include feature (9 test cases)
• Updated documentation (README, CLAUDE.md)