Bid Document Review (Lite)

AI-powered bid/tender document review. Extracts text from .docx/.doc files, cross-references bid requirements vs responses, and generates a detailed audit re...

Security Scan

The skill's files, instructions, and dependencies are consistent with a lightweight bid-document reviewer that extracts .doc/.docx text and relies on the agent/LLM for analysis; it does not request unrelated credentials or perform suspicious installation or network actions.

Purpose & Capability

Name/description match the included artifacts: SKILL.md describes text extraction and checklist-based review, and the repository contains two small Python scripts that extract text from .docx and legacy .doc files. Declared dependencies (python-docx, olefile) align with the stated purpose.

Instruction Scope

Runtime instructions only tell the agent to place bid files in a working directory, run the included extraction scripts, and perform checklist-driven analysis. The instructions do not ask the agent to read unrelated system files, environment variables, or post data to external endpoints. The agent will handle the analysis using LLM capabilities, which is expected for this skill.

Install Mechanism

No install spec is present (instruction-only plus included scripts). The skill does not download or execute third-party code at install time. It lists Python packages the user should install with pip, which is a common, low-risk approach.

Credentials

The skill requires no environment variables, credentials, or config paths. The only external requirements are standard Python libraries (python-docx, olefile) needed to parse document formats — proportional to the stated functionality.

Persistence & Privilege

The skill is not forced-always, does not request elevated privileges, and contains no code that modifies other skills or system-wide agent settings. Autonomous invocation is allowed but is the platform default and not combined with other concerning behaviors.

Assessment

This skill appears coherent and implements only text-extraction helpers plus LLM-driven review. Before installing/using: (1) verify you trust the source and inspect the small Python scripts (they only read the provided files and write output); (2) be aware that extracted text may contain sensitive procurement data — avoid sending confidential documents to external hosted models unless allowed by your data policy; (3) install python-docx and olefile from PyPI in an isolated environment (virtualenv) and run on a machine that can safely handle the documents; (4) if you need image/certificate verification or PDF support, note this lite version explicitly lacks those features and you may need the full skill or additional tooling.

Review bid/tender documents for errors, contradictions, compliance issues, and fraud indicators using AI analysis.

What it does:

• Extracts text + tables from .docx and .doc bid documents
• Cross-references procurement requirements vs bid responses
• Checks pricing consistency, qualification claims, technical parameters
• Identifies contradictions, missing information, expired certificates
• Generates a structured audit report with 🔴/🟡/🟢 risk ratings

Ideal for:

• Bid managers reviewing submissions before deadline
• Procurement officers auditing received bids
• Companies reviewing competitor bids (post-award disclosure)
• Quality assurance on your own tender responses

Quick Start

1. Place your bid documents in a working directory:

   • Procurement/tender document (the requirements)
   • Bid/response document (what was submitted)

2. Tell your agent:

   Review the bid documents in [directory]. The procurement document is [file1] 
   and the bid response is [file2].
   

3. The agent will:

   • Extract all text and tables
   • Analyze against the checklist below
   • Generate a structured report

Text Extraction

For .docx files:

python3 {baseDir}/scripts/extract_text.py input.docx output.txt

For .doc (legacy) files:

python3 {baseDir}/scripts/extract_doc_text.py input.doc output.txt

Review Checklist

1. Pricing & Commercial

• Total price within maximum limit
• Unit prices within per-item limits
• Tax rate calculations correct
• Amount in words matches figures
• No abnormally low pricing (dumping risk)
• Payment terms match requirements

2. Mandatory Requirements (★ items)

• ALL mandatory/starred parameters responded to
• Responses meet or exceed minimums
• Supporting evidence provided for each claim
• No contradictions between different sections

3. Qualification & Eligibility

• Business license valid and matching
• Required certifications in-date
• Performance track record meets minimum
• Credit/reputation checks provided
• Authorization letters (if agent/distributor)

4. Technical Response

• All technical parameters addressed
• Claims supported by test reports/certificates
• Product model matches throughout document
• Standards referenced are current (not withdrawn)
• Delivery timeline realistic vs. claimed

5. Document Integrity

• Bidder name consistent throughout
• Signatures and seals present where required
• Dates logical (no future dates, no pre-bid dates)
• Page numbering sequential
• No template placeholders left unfilled (e.g. [X], [TBD])

6. Common Red Flags

• Identical test results matching nominal values exactly (fabrication indicator)
• Contracts with missing signatures or blank fields
• Expired certificates/qualifications submitted as valid
• Third-party materials without clear authorization
• Inconsistent company names across documents

Report Format

Generate the report in Markdown with this structure:

# Bid Review Report

## Project Info
- Procurement: [name]
- Bidder: [name]
- Date: [date]

## Summary
| Category | Status | Notes |
|----------|--------|-------|
| Pricing  | ✅/⚠️/❌ | ... |
| Mandatory params | ✅/⚠️/❌ | ... |
| ...      | ...    | ...   |

## 🔴 Critical Issues (may cause disqualification)
### 1. [Issue title]
- Location: [where in document]
- Detail: [what's wrong]
- Impact: [consequence]
- Recommendation: [fix]

## 🟡 Medium Issues (affects scoring)
...

## ✅ Positive Findings
...

## Checklist Summary
[Completed checklist with pass/fail for each item]

Dependencies

• Python 3.6+
• python-docx (pip3 install python-docx)
• olefile (pip3 install olefile) — for .doc files only

Limitations (Lite Version)

This lite version covers text-based review only. It does not include:

• Image extraction and visual analysis (certificates, contracts, photos)
• Automated image fraud detection (watermarks, stock photos, expired seals)
• PDF report generation
• Image compression/optimization for API cost savings

For full image review capabilities, see the complete bid-review skill.

Built From Real Experience

This skill was developed from reviewing 3 real bid documents (1,800+ pages, 2,600+ images) in the road maintenance equipment industry. Every checklist item comes from an actual issue found in production reviews.

Issues discovered include: fake test data, stolen web images with visible watermarks, technical parameters contradicting government records by 4 tons, expired certificates submitted as valid, and 30+ pages of copied filler content.