Ace-Proposal-Respond

Respond to client proposals by converting materials to memory, creating a response strategy, and answering questions in small batches. Uses ace-context-to-memory for RAG. Same iterate-on-strategy pattern as ace-shaping.

When to Apply

User wants to respond to an RFP, Q&A, or proposal requirements
Creating a response plan from proposal documents
Answering client questions using memory (RAG)
Iterating on strategy with corrections (DO/DO NOT)

Dependency: ace-context-to-memory

Run index_memory.py --path <proposal_source> before answering questions. Use search_memory.py "<query>" when drafting answers.

Process

Setup — Convert proposal to memory; create response folder; symlink
Strategy first — Analyze documents; propose response plan; save to response/strategy.md
Answer a few questions ONLY — 3–5 per batch; use RAG; get approval
Iterate — Corrections → add DO/DO NOT to strategy; re-run or proceed

Operations

Operation	When
`create_strategy`	"Create strategy," "propose response plan," "analyze and plan"
`answer_questions`	"Answer questions," "answer a few," "next batch"
`improve_strategy`	"Correct," "fix that," "wrong"
`proceed_slice`	"Proceed," "expand," "next slice"

Build

cd skills/ace-proposal-respond
python scripts/build.py