Retainer Scope Sheet: itemize what an agent monthly covers
by PaperChaseWebb
One sheet per client that says line by line what your monthly agent retainer covers: the desks it runs and how many runs each, included hours and what happens over the limit, response times with covered hours, what is not included, what needs the client's yes, the reports they get and how changes are asked for. A checker fails every vague line ("ongoing support", "as needed", "ASAP"), an overage with no rate, a desk that sends or books with no yes row, and checks your proposals, change requests and invoices against the sheet. You approve each sheet at your own terminal.
- Price
- $29
- Difficulty
- Intermediate
- Setup
- 30 min
- Format
- .md file
retainer-scope-sheet-itemize-what-an-agent-monthly-covers.md · 230 KB · instant download after checkout
Want this built for you instead? We can build it for you.
What you end up with
Each client has a sheet with a number, a unit or a place on every line, your proposal says nothing the sheet does not, change requests wait for a written yes, and every invoice line points back to the sheet.
What you get
- 19 sections and an appendix in one .md file, with a one-line unpack for all 71 files
- The seven-part sheet: Desks covered, Hours and runs, Response times, Not included, Needs your yes, Reporting and Change requests, with fee, term and notice
- scope_check.py sheet: 18 checks, from vague words and desks with no monthly number to an overage with no rate or yes, response times with no time zone, fewer than three exclusions and a send or booking with no yes row
- scope_check.py check: 8 change request checks and 6 invoice checks against the sheet, such as a change billed before the client's written yes, overage with no recorded yes and fees at cost the sheet never names
- scope_check.py proposal: 11 checks on sales copy and quotes, such as a price that is not the sheet's fee, unlimited, 24/7, response times not on the sheet, guarantees and "no extra costs"
- scope_check.py fix: 8 checks so an agent can only add and make lines specific, never drop an exclusion, a yes row or a change step, or move your fee, limits, rates or response times
- Report and rules checks, and approve.py: it needs a real terminal and part of the sheet's hash, refuses agent sessions and piped input, and checks the client's sheet has not changed before the new one goes in
- An eval set with 14 made-up client cases, 3 proposals and 3 reports, rules and two skills for Grok Bot, Claude Code, Codex and Hermes, six prompts and templates
- The test harness: 102 unit tests and 307 planted faults you can repeat
Who it is for
Agencies and freelancers who want every monthly retainer written line by line, checked for vague items, and matched against proposals, change requests and invoices, with a person approving every sheet.
Works with
Solid chips: the agent is named in this listing's own tools list. Dashed chip: Plain Markdown, so it also pastes into other agents such as Grok Bot, Hermes, OpenClaw or Muse. We list those as compatible by format only, not as tested.
Tools needed
Cost to run: The kit is the only purchase. Everything runs locally with Python; the checker reads files only, makes no network calls and logs in to nothing.
Time to result: The worked example in 5 minutes; a first sheet for one of your clients in about 30 minutes
How it works
Unpack and test
Save the kit as pack.md in an empty folder, run the unpack line from Appendix A, then sh test_all.sh.
Expected: 71 files written; 102 tests OK; 307 of 307 planted faults caught; the eval set and rules check pass.
Watch the flow
Run sh example/run_example.sh and read it next to section 16.
Expected: A made-up client's sheet with 21 findings, the agent's first fix caught with 3, the second fix clean, a proposal with 10 findings, a first-try report with 3 and a piped approval refused.
Write one client's sheet
Edit settings.json, copy templates/SCOPE-SHEET.md into a client folder and fill in your own terms, then run python3 desk/scope_check.py check on that folder.
Expected: A list of every line that still needs a number, a unit or a place.
Check, fix and approve
Have your agent follow skills/scope-sheet-write and skills/proposal-check, run scope_check.py fix until 0 high, then approve the sheet at your terminal and send it to the client yourself.
Expected: A sheet with 0 high findings and a proposal that matches it.
Sample prompts, free to use
Read AGENTS.md. Run scope_check.py check on clients/[name] and tell me in plain words what the monthly covers today and which lines a client could read two ways. Change nothing.
Prompt 1 of 6.
Follow skills/scope-sheet-write/SKILL.md for clients/[name] and write queue/fixes/SCOPE-SHEET.md. List every number you need from me instead of making one up.
Prompt 2 of 6.
Run scope_check.py check on every client folder before invoices go out and list each V and Q code with the line it points to.
Prompt 5 of 6.
Free preview of the file
# Retainer Scope Sheet: itemize what an agent monthly covers Version 1.0, 2026-11-01. By PaperChaseWebb. For agencies and freelancers who sell a monthly agent retainer (front desk, follow-up, booking, review replies, reporting) and want one sheet per client that says line by line what the monthly covers, how many runs and hours it includes, how fast you answer, what it does not cover, what needs the client's yes and how changes are asked for, with a checker that fails every vague line. Works with Grok Bot, Claude Code, Codex and Hermes. Python 3.8 or newer, standard library only, no API keys, no logins. The checker reads files and changes nothing; it opens no network connection. Tested with 102 unit tests and 307 planted faults on Linux with Python 3.13, 3.12, 3.11, 3.9 and 3.8. All agencies, clients, people and numbers in this kit are made up. Clients ask what the monthly gets them before they sign, and again every time an invoice looks different. Agencies that list it, desk by desk with numbers, are easier to say yes to than "full coverage of your front desk". Most retainer scopes are written once, in a hurry, from the proposal. "Ongoing support." "Follow-up as needed." "Response ASAP." Overage "billed". Nothing about texts that go out without the client seeing them. Then month two: the proposal said unlimited, the invoice has a line nobody agreed to, and a small change turned into ten hours of work. An agent can write the sheet in a minute. It also writes "etc.", round
Test record
- Date
- Nov 1, 2026
- What was run
- tests/test_lib.py (8), test_sheet.py (20), test_changes.py (16), test_proposal.py (10), test_fix.py (12), test_claims.py (9), test_modes.py (8), test_set.py (3), test_mini.py (4) and test_approve.py (12, on a real pseudo terminal); tests/mutate.py with 290 planted faults across sheets, change requests and invoices, proposals, bad fixes, reports and agent rules, plus 17 against scope_mini.py; scope_check.py set on the eval set; example/run_example.sh end to end; and the unpack line with test_all.sh. Linux, Python 3.13, 3.12, 3.11, 3.9 and 3.8.
- Result
- 102 passed, 0 failed on all five versions; 307 of 307 planted faults caught and all clean baselines clean; the eval set matched 20 of 20; the example caught the bad sheet with 21 findings, the first-try fix with 3, the proposal with 10 and the first-try report with 3, and refused piped approval; the unpack line wrote every file and test_all.sh passed there.
- Not run
- A live Grok Bot, Claude Code, Codex or Hermes session. Real billing, email or CRM tools. Whether the work was done well. Legal review. macOS or Windows.
Copied from this item's dated test notes. What was not run stays in.
Worth knowing
- The agent rules were checked by the checker but not loaded inside a live Grok Bot, Claude Code, Codex or Hermes session for this release. The eval fixes were written by hand to the rules. Run scope_check.py set evals with your own agent's fixes before you rely on it.
- Not legal advice and not a contract. A sheet that passes says every line is specific and the files agree with it; it does not say the work was done well or that the terms are right for you.
- The kit reads files you save. It never logs in to billing, email or a CRM and cannot send, bill or post anything.
- The owner step assumes the agent cannot run approve.py at a terminal. Run it as a separate user; agent rules alone are not a hard boundary.
- All agencies, clients, people and numbers in the kit are made up. Tested on Linux only; macOS and Windows were not tested.
- On 2026-11-01 on Linux: 102 unit tests and 307 of 307 planted faults passed on Python 3.13, 3.12, 3.11, 3.9 and 3.8; the eval set matched 20 of 20; the unpack line wrote every file and test_all.sh passed in that folder.
Questions
How is this different from Compounding Offer Audit?
Compounding Offer Audit scores whether an offer is worth building. This kit is for after you sell a monthly: it writes down, line by line, what the monthly covers and checks your proposals, change requests and invoices against it.
Does it set my prices?
No. The fee, included runs and hours, rates and response times are yours. The checker makes sure each one is written with a number and a unit, and the fix check stops an agent from changing them.
Which agents does it work with?
Any agent that can read files and run a script. The kit has rules for AGENTS.md and CLAUDE.md, Claude Code and Codex notes, Hermes notes and a Grok Bot standing order. Those were checked by the checker, not loaded in each agent for this release.
Can my agent mark a change request approved?
No. A change request counts as approved only with the client's written yes, name and date, and an invoice line against one that is not approved fails.
Can you set it up for us?
Yes. Our Install sprint writes and checks a sheet for each of your clients and checks your proposals and invoices against them, or Managed Ops checks them for you each month. Book a call or write to [email protected].
Details
Version 1.0 · Updated Nov 1, 2026 · Markdown version of this page
License: Use inside your own team or company, including for your clients. Do not resell or republish the kit itself. Download terms
Related
Retainer Scope Sheet: the seven-part monthly scope and a ten-check sheet scan (free sample)
From Retainer Scope Sheet: the blank seven-part sheet, ten rules to paste into your agent, and a read-only script that runs ten checks on one scope sheet, from a missing section and vague lines like "as needed" to a desk with no monthly number, an overage with no rule, a response time with no time zone, too few exclusions, a send or booking with no yes row and a result promise.
Compounding Offer Audit: score an offer on six tests, find its weakest point, change one thing
Score each thing you sell from 0 to 3 on six tests: recurring, productized, systemized, owned demand, compounding, and whether growing it would make the owner's week better. Every score needs a checkable fact and its source, or the word unknown and the fact you are missing; no 3 on the owner's say-so, and no total. The audit names the weakest test, treats a low life score as a veto on growth, and proposes one change with a success test and a date. Every change waits for the owner's yes. Fifteen read-only checks run on the folder.
Price, pitch and follow up: prompts for solo founders selling services
15 prompts for pricing a service, writing a one-page offer and proposal, handling objections and following up on a schedule, plus a read-only checker that flags a price under your own minimum, numbers that are not in your notes and which follow-up is due for each lead.
Invoice Chase Desk: AI-drafted payment reminders, checked against your ledger, approved by the owner
Let an agent draft your overdue invoice reminders without chasing a client who already paid. A planner reads your ledger and decides who gets a reminder today, and at which step. A checker refuses drafts with the wrong balance or dates, a fee your terms do not include, threats or pressure, new bank details, lookalike payment links, other clients' invoices or claims nobody can back up. Disputes, money trouble and odd replies go to the owner, and before anything is recorded as sent the desk checks the ledger again.
Consultant and Agency Skills Pack: discovery brief, scope, change request, client update and project close
Five installable agent skills in the Agent Skills format that carry one client engagement from first call to handoff, each drafting only from your notes with [ASK] for missing facts and never sending anything, plus skill_check.py, ten read-only checks on any skills folder against the spec and draft-only rules.
Ask-first rules for AI agents: hard nos, approval cards and a test set
Paste-in rules that make an agent ask before it sends, spends, posts or deletes, define what a real yes is, and stop retries after a no, plus 12 test scenarios to check it.
Solo-founder agent ops playbook: run your business with agents without losing control
Decide which jobs go to agents and how much power each gets, write one house rules file with five numbered stops, then run a read-only checker that compares your agent register with the action log, budget and incidents every week.