pae/crimson_leaf
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

index: add proposal {task.id} to proposal index

Browse Source
This commit is contained in:
PAE
2026-05-01 20:32:46 +00:00
parent 0ffbd2d3b8
commit 6331aabae6
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
deliverables/proposals/index.md
View File

@@ -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.
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.