Llm Tooling Python Packages

agentseam

The primitives layer for every coding agent — one handler API over per-agent hooks, instruction files, plugin packaging, and config.

hermes-prime-mcp

Drift-prevention session-init convention card for fresh Claude Code sessions. Injects a self-contained card so a new session opens with its grounding triggers, calibration rules, and tool-map in scope — instead of re-deriving them at minute 30. Per-project, marker-anchored, idempotent, reversible. Bash installer plus MCP server.

lightningrod-ai

Better AI Predictions ⚡

toolahead

Speculative execution for AI coding agents — predicts upcoming Codex and Claude Code tool calls, safely pre-runs them, and replays validated results instantly. Pre-warms dev servers while the model reasons. Local-only, MIT.

quale

Structural code analysis for LLMs, humans, and CI -- no parsers, zero config.

codecompass-context

Evidence-backed context for AI coding agents. CodeCompass maps pinned dependencies, project relationships and technical context into a queryable graph, then evaluates whether that context actually helps agents. Pre-release; v1 is being validated on real projects.

self-improving-loop

Rollback-first reliability layer for AI agents: trace failures, apply guarded changes, and rollback on regression.

skillprune

Audit your installed agent skills against real usage. Finds what never fires, what collides, and what it costs you every turn. Stdlib only.

finasys

finasys: Fast financial data processing, feature engineering & AI agent toolkit

wulong

A governance and policy layer that wraps Claude Code with review gates, audit trails, and plain-English rules for multi-agent engineering.