squawkit 0.2.0

Semi-QUalified Agent Wingman Kit — the stateful intelligence + MCP server layer for fledgling-equipped agents.

squawkit

Semi-QUalified Agent Wingman Kit. The stateful intelligence + MCP server layer for fledgling-equipped agents.

squawkit wraps fledgling's SQL macros (via pluckit — pip install ast-pluckit) with smart defaults, token-aware output, session caching, compound workflows, an MCP server, prompt templates, and live resources.

Install

pip install squawkit

Run

squawkit

Starts the FastMCP server on stdio. Connect it to Claude Code, Cursor, or any MCP-compatible client.

What agents get

• 25+ tools — code search, AST analysis, doc browsing, git history, diagnostics
• 4 compound workflows — explore, investigate, review, search
• 3 prompt templates — pre-loaded with live project data
• 5 resources — always-on project overview, docs, git status, session log
• Smart defaults — infers language, doc layout, main branch automatically
• Token-aware output — truncation with head+tail hints, automatic bypass

Architecture

squawkit → ast-pluckit → fledgling-python → fledgling (SQL) → DuckDB extensions

Note: pluckit is published on PyPI as ast-pluckit. The Python import name is still pluckit.

squawkit is the opinionated top layer. It adds session state, MCP protocol, and intelligence heuristics on top of the stateless query layers below it.

Documentation

Full docs at squawkit.readthedocs.io.

License

Apache 2.0