OTel-native typed primitives for LLM cost attribution and telemetry — published on PyPI.
High-assurance in-memory Tree-Sitter AST context firewall and pruning MCP server for coding agents (-72.4% token mass).
PyPI-distributable LLM control plane: gateway choke point, cost attribution, OTel instrumentation, and offline reporting as an inspectable engineering artifact.
FinOps cost attribution for AI code agents: maps Kiro, Cursor, and Claude Code spend to git commits to reveal per-task cost, waste patterns, and agent ROI signals.
Tamper-evident, stranger-verifiable receipts for LLM cost-savings — anyone can recompute your caching/routing savings math offline, no trust in your dashboard required. Pure stdlib, zero-dependency.
Estimate LLM API costs from token usage and request volume using configurable provider pricing.
Evidence-based quality gate for LLM deployments: evaluates telemetry against latency, cost, and error policies to produce auditable go/no-go release decisions for CI/CD.
Pre-dispatch policy evaluation and cost attribution for LLM inference, built on llmscope.
Open-source LLM observability, FinOps, and governance platform - normalize token usage, cost, latency, errors, retries, quotas, and billing across OpenAI, Anthropic, Gemini, Azure OpenAI, and AWS Bedrock. OpenTelemetry-native, Prometheus-ready, self-hosted. Apache-2.0.
Claude skill for code-agent execution policy: scope control, credit burn reduction, and cheap context recovery for spec-driven implementation workflows.
To associate your repository with the llm-finops topic, visit your repo's landing page and select "manage topics."