# Bid Watch Sheet: score this week's notices before anyone replies

> A free kit for small firms that check government contract notices every week. Write a profile once: your NAICS codes, the set-asides you hold, your states, the shortest runway you will work with, keywords and exclude words. Each week, put the notices in a CSV from the search you already use. Five gates pass the clear misses with a reason, the rest get a fit score out of 100, and nothing is drafted until the owner records pursue or pass at their own terminal.

- URL: https://paperchasewebb.ai/bots/marketplace/items/bid-watch-sheet-score-this-weeks-notices-before-anyone-replies/
- Type: Prompt pack
- Category: Solo business
- By: PaperChaseWebb
- Works with: Claude Code, Codex, Grok Bot, Hermes (named in the listing); One Python script, a profile template, a fictional firm with eight fictional notices, and agent rules for each harness. Tested on Linux
- Price: Free USD
- Difficulty: Beginner
- Setup time: 15 min
- Time to result: First watch list in about 10 minutes
- Format: Markdown (.md) file, bid-watch-sheet-score-this-weeks-notices-before-anyone-replies.md
- Version: 1.0, updated 2026-11-03
- License: Personal and commercial use.

## What you end up with

A short, ranked list each week, and no reply to a buyer the owner did not choose.

## What you get

- The method: apply your hard facts first, score what is left, and let the owner choose
- bidwatch.py: 5 profile checks, 5 gates (NAICS, set-aside, runway, excluded words, place) and a fit score with the reason shown
- watch.md each week: ranked notices and the passed ones with the gate that stopped them
- Pursue or pass refused inside an agent session, without a terminal and from anyone but the owner; decisions flag themselves when a notice is amended
- AGENTS.md, CLAUDE.md, a Grok Bot standing order and a Hermes skill, a fictional example, and 26 tests

## Who it is for

Small firms and solo contractors who check notices every week and want the owner to choose before anyone replies.

## Tools needed

- Python 3.8 or newer: bidwatch.py: profile check, 5 gates, a fit score, watch list, owner pursue or pass, and status with due-soon and changed flags; standard library only; reads a file and never goes online
- Grok Bot, Claude Code, Codex or Hermes (optional): fills the notices file, runs the scan and drafts only for notices marked pursue under AGENTS.md; never decides, contacts a buyer or submits anything
- You need: Python 3.8 or newer
- You need: A weekly list of notices from the search you already use

Cost to run: Free. Python is free.

## How it works

1. **Run the example.** Unpack the file and run python3 bidwatch.py scan example/profile.json example/notices.csv out/ --today 2026-11-03. Expected: 3 notices on the watch list; 5 passed, one per gate.
2. **Write your profile.** Copy templates/profile.json and fill in your codes, the set-asides you hold, your states and your runway. Expected: bidwatch.py check shows profile OK.
3. **Scan a real week.** Put this week's notices in notices.csv and run scan. Expected: watch.md ranked, with a reason for every score and every pass.
4. **Decide at your own terminal.** Read the full notice, then run decide with pursue or pass. Drafting starts only on pursue. Expected: status shows each decision, DUE SOON items, and CHANGED if a notice is amended.

## Sample prompt: Monday bid watch (Grok Bot, Claude Code, Codex, Hermes)

```
Follow AGENTS.md. Put this week's notices in notices.csv exactly as listed, run bidwatch.py scan and status, and send the owner watch.md with anything due within 14 days. Draft only for notices marked pursue. Do not decide, contact a buyer or submit anything.
```

One prompt.

## Free preview of bid-watch-sheet-score-this-weeks-notices-before-anyone-replies.md

````markdown
# Bid Watch Sheet

**Score this week's contract notices against your own profile. Clear misses get passed with a reason, the rest get ranked, and nobody replies to a buyer until you say pursue.**

Version 1.0, 2026-11-03. By PaperChaseWebb Inc. Free. Works with Grok Bot, Claude Code, Codex and Hermes, or by hand. Python 3.8 or newer, no installs.

## The problem

A weekly search turns up dozens of notices. Most are wrong for you on one hard fact: the code is not yours, the set-aside is one you do not hold, it closes in four days, or it is in a state you do not work in. Reading each one to find that out eats the week, and an eager agent will happily draft replies to all of them.

This kit applies your hard facts first, scores what is left, and stops at a list. The owner says pursue or pass on each notice, and only then does drafting start.

## What you get out of it

- watch.md: the notices that cleared every gate, ranked, with the reason for each score, and the ones passed with the gate that stopped them.
- A pursue or pass record per notice, made by the owner, that flags itself when the notice is amended.
- A Monday routine your agent can run without being able to answer for you.

## Who it is for

Small firms and solo contractors who check government opportunities every week, and the people who run that check for them.

Not legal or contracting advice. You confirm your own eligibility, and you read the full notice before you decide.

## The profile

| Field | What to write 
````

## Test record

- Date: Nov 3, 2026
- What was run: tests/test_bidwatch.py (26 tests) unpacked from the pack file and run fresh on Linux, Python 3.8, 3.9, 3.11, 3.12 and 3.13; check, scan, decide and status on the fictional example.
- Result: 26 passed, 0 failed on all five versions; the example puts 3 notices on the watch list and passes 5, one per gate; decisions refused in agent sessions, without a terminal, for the wrong name and without the typed choice; changed notices flagged; the profile and notices are never modified.
- Not run: macOS or Windows. Real notice exports. A live Grok Bot, Claude Code, Codex or Hermes session.

Copied from this item's dated test notes. What was not run stays in.

## Worth knowing

- It reads the fields you give it. A wrong due date or code in the CSV gives a wrong answer.
- Keywords match the title only. A score sorts the list; it is not a reason to bid.
- On 2026-11-03: 26 tests passed on Python 3.8, 3.9, 3.11, 3.12 and 3.13 on Linux, run from the files unpacked out of the pack. Agent rules guide; they were not loaded inside each harness for this release.
- Not legal or contracting advice. You confirm your own eligibility.

## Questions

**Does it search for notices?** No. It reads a CSV you or your agent fill from the search you already use. That keeps it offline and safe to hand to an agent.

**Can my agent reply to a buyer?** No. Only the owner records pursue or pass, and the agent rules forbid contacting a buyer or submitting anything.

**Why pass a notice outside my codes instead of scoring it low?** A hard miss is not worth reading. Every pass shows the gate, so you can fix the profile if a gate is wrong.

**What should I pair it with?** GovCon Opportunity Triage Kit ($39) takes one notice you marked pursue through a full bid or no-bid, and the Eligibility and Compliance Gate Checklist ($29) checks the rules before you commit.

## How to get it

Free sample. Download it directly at https://paperchasewebb.ai/bots/marketplace/items/bid-watch-sheet-score-this-weeks-notices-before-anyone-replies/, no signup.

Seller: PaperChaseWebb Inc., mgmt@paperchasewebb.com