diff --git a/deliverables/proposals/index.md b/deliverables/proposals/index.md index 1e8d887..d316b25 100644 --- a/deliverables/proposals/index.md +++ b/deliverables/proposals/index.md @@ -66,7 +66,7 @@ Date: 2026-04-29 Status: AWAITING DAVID'S APPROVAL Summary: Proposal for the Foreman Probe project to model probe tasks created by the Foreman to benchmark and evaluate LLM capabilities. This fills the gap in internal evaluation by providing a focused framework for testing LLM performance in Foreman-specific scenarios. It differs from prior proposals by emphasizing the direct application of these tasks in operational contexts, integrating both technical metrics and practical workflows for more robust validation. -### Crimson Leaf Holdings -- Task bf9a4a39-de0e-402b-af3d-c4da508c810c +### Crimson Leaf Holdings -- Task ba47113f-0cea-4d9e-bf7a-f847408ab3a2 Date: 2026-04-29 Status: AWAITING DAVID'S APPROVAL -Summary: Proposal for the Foreman Probe project leveraging construction industry search results on Foreman planning, estimating software, and proposal workflows to create authentic probe tasks that simulate real-world contractor operations. This fills the gap in domain-specific benchmarking by grounding LLM evaluation in practical construction management practices like look-ahead planning and bid proposals. It differs from prior proposals by directly incorporating external industry resources and tools such as Contractor Foreman demos to model realistic Foreman task generation in construction contexts. \ No newline at end of file +Summary: This proposal is for the company to consider the creation of agentic probe tasks that model tasks created by the Foreman. The goal is to evaluate LLM performance within Foreman-specific operational contexts. This proposal differs from prior ones by focusing on agentic aspects of LLM reasoning within controlled Foreman-generated task environments. \ No newline at end of file