[Trend Scout] srclight/srclight

[github-actions]

🔭 Trend Scout: srclight/srclight

Auto-generated by trend-scout.py — review and edit as needed.

📌 What problem it solves

Deep code indexing MCP server for AI agents. 25 tools: hybrid FTS5 + embedding search, call graphs, git blame/hotspots, build system analysis. Multi-repo workspaces, GPU-accelerated semantic search, 10 languages via tree-sitter. Fully local, zero cloud dependencies.

📅 Timeline

Field	Value
Created	2026-02-23
Last pushed	2026-04-27
Stars	33
Forks	7
Open issues	3
License	MIT
Language	Python
Topics	code-indexing, code-intelligence, code-search, developer-tools, embeddings, fts5, mcp, mcp-server, ollama, python, semantic-search, sqlite, tree-sitter

✅ Strengths

• Growing community (33 ⭐)
• Well-tagged: code-indexing, code-intelligence, code-search, developer-tools, embeddings, fts5
• Primary language: Python
• Actively maintained (pushed within 30 days)

⚠️ Weaknesses / Risks

• No significant risks identified from available metadata

💡 What this repo can learn

• Claude Code session patterns: this repo's Claude Code integration approach could improve claude-adapter.py's JSONL parsing — e.g., handling new session event types or extracting richer metadata from Claude Code tool-use blocks
• Hybrid FTS+semantic retrieval: combining keyword and embedding-based search could improve recall in query-session.py / briefing.py — e.g., a query for 'docker networking' would also surface entries tagged 'container' or 'network_mode' even without exact term overlap
• CLI verb patterns: a clear add/search/update/delete verb model (like memory-tool add / search / dream) could streamline the UX of query-session.py and learn.py, making them easier to invoke from hooks or scripts
• Offline-first design: this repo's no-cloud/no-server posture directly mirrors our local-SQLite approach — could validate that knowledge.db workflows never require external API calls even when semantic search is enabled
• Git hook / workflow patterns: hook design from this repo could strengthen the hooks/ enforcement chain (e.g., auto-briefing, commit guards, learn reminders)

README excerpt

<!-- mcp-name: io.github.srclight/srclight -->

# Srclight

[![PyPI](https://img.shields.io/pypi/v/srclight)](https://pypi.org/project/srclight/)
[![License](https://img.shields.io/github/license/srclight/srclight)](https://github.com/srclight/srclight/blob/master/LICENSE)
[![Python](https://img.shields.io/pypi/pyversions/srclight)](https://pypi.org/project/srclight/)

**Deep code indexing for AI agents.** SQLite FTS5 + tree-sitter + embeddings + MCP.

Srclight builds a rich, searchable index of your codebase that AI coding agents can query instantly — replacing dozens of grep/glob calls with precise, structured lookups. It is the most comprehensive code intelligence MCP server available: 42 tools covering symbol search, relationship graphs, community detection, impact analysis, git change intelligence, semantic search, build system awareness, and document extraction — capabilities no other single MCP server combines. Fully local and private: your code never leaves your machine.

## Why?

AI coding agents (Claude Code, Cursor, etc.) spend **40-60% of their tokens on orientation** — searching for files, reading code to understand structure, hunting for callers and callees. Srclight eliminates this waste.

| Without Srclight | With Srclight |
|---|---|
| 8-12 grep rounds to find callers | `get_callers("lookup")` — one call |
| Read 5 files to understand module | `codebase_map()` — instant overview |
| "Find code that does X" → 20 greps | `semantic_search("dictionary lookup")` —

*(truncated)*

---

Scouted on 2026-05-02 · View on GitHub

[magicpro97]

Closing this auto-generated Trend Scout discovery issue to keep the tracker focused on actionable product/runtime defects.

This item is a research lead rather than a verified bug in the repo. The current release wave has closed the concrete browse/tooling issues, so I’m triaging this discovery out of the active issue queue.