v0.4.1 — PyPI metadata refresh for category-search discovery

What's Changed

Metadata-only release. No code, no test, no protocol changes — wire format and public API are identical to 0.4.0. pip install --upgrade runcycles is safe; no migration needed.

Why

Diagnostic on Python-side adoption found the biggest gap was discovery: runcycles wasn't ranking for natural Python search queries ("Python AI agent budget control", "LLM cost enforcement", etc.) while named competitor packages were. The brand-led PyPI description and missing keywords were contributors.

Three changes to pyproject.toml

1. Description rewritten to lead with the literal category-search phrase:

   • Old: "Python client for the Cycles budget-management protocol"
   • New: "Python AI agent budget control — enforce LLM cost limits, tool permissions, and multi-tenant policies before agent actions execute."

2. Keywords expanded 12 → 21, organized into:

   • Category-search: ai-agent, agent-budget, agent-governance, budget-control, cost-control, cost-enforcement, spending-limit, llm-cost, runtime-authority, action-control, multi-tenant
   • Framework targeting: langchain, langgraph, crewai, autogen, openai-agents, mcp, openai, anthropic
   • Brand: cycles, runcycles

3. Classifier added: Topic :: Scientific/Engineering :: Artificial Intelligence — standard PyPI category for AI/ML packages.

Companion changes (separate)

• GitHub topics on this repo: governance dropped, mcp added (now mirrors cycles-client-typescript).
• runcycles/docs: Python framework integration guide titles retitled for category-keyword leading (PR runcycles/cycles-docs#568).

• chore(seo): retarget PyPI metadata for category-search discovery by @amavashev in #56

Test coverage: 100% across all 13 modules (no test additions; no behavior change).

Full Changelog: v0.4.0...v0.4.1