Bid Proposal Multi-Agent Generation
Paste the following prompt into your AI chat to install this skill:
Please install @user_c542ade9/bid-proposal-zh following https://skillhub.cn/install/skillhub.md.
About this skill
Problem
Generating bid documents is often hard because tender clauses are easy to miss, chapter drafts are assembled manually, and AI-generated content may contain unsupported facts, numbers, or case references. Treating a bid as a generic long document can lead to incomplete compliance matrices, policy errors, technical deviations, inconsistent business terms, and format failures.
How It Works
The skill runs a six-stage pipeline:
- Dispatcher parses the tender and outputs tender elements, a policy-compliance report, chapter briefs, and a draft compliance matrix
- Parallel chapter agents write sections with knowledge-base-first retrieval from IMA, marking unsupported content for review
- Compiler merges the full proposal, compliance matrix, and technical/business deviation tables
- QA applies five hard gates and rejects work if any gate fails
- Five reviewers score completeness, technical depth, business competitiveness, language quality, and scoring strategy
- Revision loop merges P0/P1/P2 findings and iterates up to three rounds before delivery
Boundaries
Best for Chinese bid drafting when the tender file, qualifications, cases, or technical materials are available. If IMA or Lixiang knowledge bases are unavailable, AI-generated coverage increases and fields such as certificate numbers, pricing, and contract references still need human confirmation.
Use Cases
- After receiving a tender, extract qualification, technical, commercial, and scoring requirements into a draft compliance matrix.
- Before drafting technical chapters, retrieve company cases, qualifications, and technical docs from IMA and mark unmatched content.
- After the first draft, run five hard QA gates and revise issues by P0/P1/P2 severity across up to three iterations.
- After compiling the bid, produce technical and commercial deviation tables plus a full response matrix for review.
Best For
- Pre-sales engineers preparing bid documents need to split tender requirements into reviewable technical chapters.
- Bid coordinators need to check compliance matrices, deviation tables, and hard format gates to reduce rejection risk.
- Solution architects need to draft chapters grounded in qualifications, cases, and technical materials from a knowledge base.
- Project managers need to track five-dimension review reports, P0/P1/P2 issues, and iteration rounds before deciding delivery.
Related Skills
Collects daily conversations, task logs, and errors, then runs a nine-module review to surface blind spots, generate actions, and update memory.
Local long-term memory for AI agents that appends key facts and daily logs, supports cross-session context and search recall, and auto-archives old diaries to gzip files.
Run a grilling session to interact with or test AI agents.
A systematic prompt optimization skill that refines prompts using a four-step distillation framework (diagnose, structure, think, compress) and methodologies from four prompting masters.