Agency AI Ops Workbook: the Install sprint, self-serve
by PaperChaseWebb
A two-week workbook for owner-led agencies and small businesses: install one AI agent workflow yourself with the same steps as our $5,000 Install sprint. Discovery and mapping, picking the first workflow from your Self-Check score, an agent card with an owner and a tested pause, data access and a test set, an approval gate register, handoff packets, cost limits, tool vetting, a baseline and stop rule, a dry run, a go-live checklist and a 30-day review. Ten worksheets, 13 prompts with harness notes.
- Price
- $39
- Difficulty
- Intermediate
- Setup
- 1 hr 30 min
- Format
- .md file
agency-ai-ops-workbook-install-sprint-self-serve.md · 43 KB · instant download after checkout
Want this built for you instead? We can build it for you.
What you end up with
In about two weeks one workflow runs with an AI agent that drafts, a named owner who answers for it, gates enforced in the tools, a cost limit and a stop rule, and a 30-day review that tells you to keep, fix, pause or remove it.
What you get
- 21 sections in one .md file
- A day-by-day two-week plan with an approval step every day
- A rules block to paste into every agent session
- Ten worksheets: workflow inventory, one-page map, pick matrix, agent card, source list and test set, gate register, handoff packet, cost sheet, tool intake, baseline and scorecard
- Four action classes and an approval request format
- A dry run with a pass rule
- A go-live checklist and a 30-day review sheet
- 13 copy-paste prompts for Grok Bot, Claude Code, Codex and Hermes, with harness notes
- A failure table
- A worked example, labeled as illustrative
- When to hire the $5,000 Install sprint instead, and what Manage covers after
- A competitor teardown and credited sources
Who it is for
Owners and operations leads who want one AI agent workflow running properly, with gates, an owner and a measure, and would rather do the install themselves than hire it out.
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 workbook is the only purchase. Uses the agent harness and accounts you already have.
Time to result: About two weeks to go-live for one workflow, then a 30-day review
How it works
Score and map
Start from the free Self-Check. Days 1 and 2: Worksheets A and B with prompts 1 and 2.
Expected: A signed one-page map for three or four candidates.
Pick one workflow
Day 3: the pick matrix and pick rules. Nothing that touches money or reaches a client unread goes first.
Expected: One workflow and a runner-up.
Set authority, data and gates
Days 4 to 6: agent card with an owner and tested pause, source list and test set, gate register with every gate tested on purpose.
Expected: Signed cards and a gate register each approver accepted.
Handoffs, cost, tools, measure
Days 7 to 10: handoff packets with a confirm rule, a cost limit set in the tool, tool intake, a baseline and a signed stop rule.
Expected: Worksheets G to J filled and signed.
Dry run, go live, review
Days 11 to 14: ten test cases in draft mode, the go-live checklist, supervised go-live. Day 44: the 30-day review.
Expected: A keep, fix, pause or remove decision made from numbers.
Sample prompts, free to use
Follow the install rules. Here is the agent card and the signed workflow map. Walk every step and list every action that is reversible or external. For each, fill a row of the gate register: approver, answer-within time, what the agent does while it waits, and how the tool could enforce the gate. List any step where the map and the card disagree. Do not mark anything as tested; I do that.
Prompt 6 of 13. The full text and the register are in section 7.
Check each line of the go-live checklist against the worksheets. Answer yes, no or unclear and quote the evidence. List every no and unclear first. Tick a line only if the worksheet shows the thing was done and signed.
Prompt 12 of 13. The agent checks; the owner and approver sign.
Free preview of the file
# Agency AI Ops Workbook: the Install sprint, self-serve A workbook by PaperChaseWebb for owner-led agencies and small businesses that want one AI agent workflow running properly, with gates, an owner and a measure, and would rather do the install themselves. It follows the same steps as our $5,000 Install sprint, laid out as a two-week plan you can run with your own team: discovery and workflow mapping, picking the first workflow, an authority map with an owner per agent, data access, approval gates, handoffs, cost limits, security and vetting of tools, measurement and a stop rule, a dry run in draft mode, a go-live checklist and a 30-day review. Ten worksheets, 13 copy-paste prompts for Grok Bot, Claude Code, Codex and Hermes with harness notes, a failure table, a worked example and an honest section on when to hire the sprint instead. Plain .md. Version 1.0. Last tested 2026-10-27 (see "How the prompts were tested" in section 17). Works with Grok Bot, Claude Code, Codex and Hermes, and pastes into Claude or ChatGPT. Most AI agency playbooks are about selling AI services: niches, outreach scripts, pricing calculators. Very few cover the part that decides whether the work holds up after launch, which is who may approve what, who answers for each agent, how work passes between agents and people, what each run costs and when to switch the agent off. That part is what this workbook is for. It is narrow on purpose: one workflow, about two weeks, done properly. Then the next o
Worth knowing
- Not legal, tax or compliance advice. If your work is regulated, have the gate register and source list reviewed by someone qualified.
- The worked example is invented to show how the worksheets fit together. It is not a client and its numbers are not results.
- Nothing in this workbook promises time saved, revenue or any specific result. The scorecard is how you find out.
- Prompts were dry-run against the worked example on 2026-10-27. Tools change; run the day 6 gate tests in your own setup.
Questions
How is this different from the Install sprint?
Same steps, different people. In the workbook your team does the worksheets and answers for the result. In the $5,000 Install sprint we set it up with you. Section 18 lists when hiring the sprint is the better call.
Do I need to score my workflow first?
Yes. Run the free Agent Readiness Self-Check first. The workbook assumes Draft-only pilot or better and uses your skeptical score to pick the first workflow.
Will the agent send or spend anything?
No. The rules block keeps agents to reading and drafting, every external action is a gate with a named approver, and the workbook has you remove send, post, spend and delete permissions in the tools wherever they allow it.
Which tools does it work with?
Grok Bot, Claude Code, Codex and Hermes, with harness notes for each. Every prompt also works pasted into Claude or ChatGPT.
What happens after the first workflow?
The 30-day review decides keep, fix, pause or remove. After a keep, you rescore the next workflow and run the workbook again, or move to Manage ($5,000/mo) on the paperchasewebb.ai homepage.
Details
Version 1.0 · Updated Oct 27, 2026 · Markdown version of this page
License: Personal and commercial use inside your own business, including installs you run for your own clients. Do not resell, republish or share this file. Download terms
Related
The approval gate register and the gate audit prompt (free sample)
From the Agency AI Ops Workbook: four action classes, the gate register with how to fill each column, the approval request an agent writes instead of acting, and a gate audit prompt for Grok Bot, Claude Code, Codex or Hermes.
Grok Bot Agency Pack: 5 client-ready templates
Five agent desks for the work an agency repeats for every client: Client Intake, Scope and Proposal, Kickoff and Access, Weekly Client Report, and a separate Delivery Reviewer. Each comes with a persona, an authority map, a standing order or event routine, a handoff packet and five test cases. Plus the agency rules block, a client isolation rule, and two full packs: Grok Bot Templates and the Agency Open-Loop System.
Agent Readiness Self-Check: score one workflow before you hand it to an AI agent
A free self-check for agencies and small businesses. Score one workflow across eight areas (workflow mapped, data access, approval gates, owner per agent, handoffs, cost tracking, security and vetting, measurement) with an evidence rule and three gate rules, read your band and what to do next, plus an interview prompt for Grok Bot, Claude Code, Codex or Hermes, a score sheet and a worked example.
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.
Cross-Harness Relay: Grok Bot, Claude Code, Codex and Hermes on one job
A bot template and teaching guide for running more than one agent harness on the same work. Six seat cards (front desk, planner, builder, reviewer, tester, night shift), a RELAY.md seat map, a relay packet for every handoff, four loop shapes, routine-or-judgment routing with a fallback when limits run tight, a preflight check, the front desk and night shift split with a proof footer, shared memory rules, eight relay drills, a rules block and 14 prompts.
Agent Cost Levers: cut cost per solved task without cutting quality
A Markdown pack that measures what your coding agent spends per finished task, then works five levers one at a time: thinking depth, context curation, direct tool routing, cache and reuse, and a verified stop. Includes a baseline sheet, a rules block and 10 prompts.
Vet Before You Install: ten checks, a sandbox recipe and a skill for third-party agent code
A Markdown pack for deciding whether a skill, plugin, mod, MCP server or repo is safe to let into your agent setup: read it first, pin and vendor it, trial it in a sandbox with a planted fake key, then write the decision down.
Governed Agent Teams: subagent personas, authority maps and handoffs
A Markdown pack for running a team of coding and research agents with clear limits: 7 subagent personas for Claude Code with Codex AGENTS.md and Hermes skill versions, a Desk Authority Map, three role risk classes, a handoff packet the receiver verifies, an evidence standard, a decision ledger and a reach inventory.