squackit

Semi-QUalified Agent Companion Kit. A code intelligence toolkit that works two ways: as an MCP server for AI agents, and as a CLI for humans.

Built on pluckit (CSS selectors over ASTs) and fledgling (SQL macros over DuckDB).

Install

pip install squackit

CLI usage

# Run any tool from the shell
squackit tool list
squackit tool view "src/**/*.py" ".fn#main"
squackit tool find_names "**/*.py" ".class"
squackit tool complexity "src/**/*.py" ".fn"

# JSON output for scripting
squackit --json tool find "src/**/*.py" ".fn"

# Pluckit chain queries
squackit pluck "src/**/*.py" find .fn containing cache names
squackit pluck "src/api.py" find .class#Handler children find .fn names

# Start the MCP server
squackit mcp serve

Tool names resolve in three conventions: find_names, find-names, FindNames. Tab completion is built in (eval "$(_SQUACKIT_COMPLETE=bash_source squackit)").

What agents get via MCP

• ~35 tools — CSS-selector code queries (via pluckit), file I/O, git history, doc navigation, conversation analytics, and diagnostics. The exact set is dynamic (fledgling macros + pluckit tools + pluckins); a full install exposes ~40 MCP tools including the workflows below.
• 4 compound workflows — explore, investigate, review, search
• 3 prompt templates — pre-loaded with live project data
• 5 resources — project overview, docs, git status, session log, diagnostics
• pluck tool — full pluckit chain grammar for multi-step AST queries, with mutation safety (blocked by default, opt-in via allow_mutations=true)
• Smart defaults — infers language, doc layout, main branch automatically
• Token-aware output — configurable truncation with max_results/max_lines

Architecture

Layer 4   Consumers (Claude Code, agents, IDE extensions)
              │
Layer 3b  squackit ─── CLI + MCP server + intelligence
              │        smart defaults, caching, truncation,
              │        workflows, pluckit tools, mutation safety
              │
Layer 3a  pluckit ──── fluent Python API (CSS selectors over ASTs)
              │
Layer 2   fledgling ── SQL macros + Python bundler
              │
Layer 0   DuckDB extensions (sitting_duck, markdown, duck_tails, read_lines)

Pluckit tools take priority over fledgling equivalents when both provide a capability (e.g., pluckit's find supersedes fledgling's find_definitions).

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

Documentation

Full docs at squackit.readthedocs.io.

License

Apache 2.0