Current Stack Roster
Slot pages say what to select for. This page says what we picked.
- Candidates under evaluation, not endorsements; this page is a Draft
- Each row names a product in a Stack slot
- Each description is the vendor’s own, as of 2026-10, with a source link
- Change a row through a Discussion, like any stack change (FAQ)
Agent Architecture
Section titled “Agent Architecture”Slot: Agent Architecture & Orchestration
- None recorded
Cost Management
Section titled “Cost Management”Slot: Cost Management
- None recorded
Evaluation
Section titled “Evaluation”Slot: Evaluation & Benchmarking
- None recorded
Harness
Section titled “Harness”Slot: Harness Selection
| Harness | Vendor description (as of 2026-10) |
|---|---|
| Claude Code | Agentic coding tool; terminal, IDE, desktop app, and browser (source) |
| Cursor | AI coding agent (source) |
| Aider | AI pair programming in your terminal (source) |
| Custom | Built and maintained in-house |
Memory
Section titled “Memory”Slot: Memory & Context Management
- None recorded
Models
Section titled “Models”Slot: Model Selection & Routing
- None recorded
Observability
Section titled “Observability”Slot: Observability & Logging
| Tool | Vendor description (as of 2026-10) |
|---|---|
| Langfuse | Open source agent evals and observability (source) |
| Datadog | Agent observability (source) |
| Custom logging | Built in-house for domain-specific needs |
Repository Context
Section titled “Repository Context”Slot: Repository Context
- None recorded
Skills
Section titled “Skills”Slot: Skills & Prompts
- None recorded
Tool Integration
Section titled “Tool Integration”Slot: Tool Integration
- None recorded