Guide · Agent skill
How to use the review-forecast agent skill for sales forecasts
Updated · written with AI from the project's own documentation and checked against our catalogue data
The review-forecast skill is a set of instructions that lets an AI agent review a sales pipeline the way a forecast call would. It builds a rollup of open deals by forecast category and stage, flags risky or stale deals, and suggests changes to categories and follow-up actions. This guide explains what an agent skill is, how to install this one, how to ask for a review, and what it needs from you before it runs.
Sales Forecast ReviewXiaomiMiMo/MiMo-Code · MIT Download the skill Quick start
- 1Install the skill folder. Download the ZIP from the page, then unzip the review-forecast folder into ~/.claude/skills/ or the project's .claude/skills/ folder, or upload it under Settings → Capabilities → Skills in the Claude apps.
- 2Prepare your forecast data. Have a CRM connection, a CRM report or export, or a pasted pipeline snapshot ready. Without one of these the skill will stop rather than infer a forecast.
- 3Ask for a review in plain language. State the review you want, such as a forecast-call preparation or deal-risk check, and name the scope if you can.
- 4Confirm scope and conventions. Answer the skill's intake questions about the owner book, team, report or deal set, plus the posture standard, category convention and amount basis it proposes.
- 5Review the rollup and risk posture. Read the rollup by category and stage, then the flagged deals and the recommended category changes.
- 6Check evidence gaps and act on follow-ups. Note which evidence is missing, then carry out any follow-up actions yourself, since the skill does not change CRM records.
What the review-forecast skill makes the agent do
The skill is written for a sales manager or RevOps person preparing a weekly or quarterly forecast call, and for sellers who want a risk check on their own book before a manager review. It turns CRM or exported pipeline truth into a manager-ready readout: a forecast rollup, a risk posture, recommended changes to forecast categories, evidence gaps, and follow-up actions.
The skill is built around the idea that the forecast must rest on data. It does not infer a forecast from anecdotes. It asks you to confirm the scope (for example an owner book, a team, a named report or a set of deals) and the data source before it produces a posture review.
- Rollup of open deals by forecast category and stage, with deal count, primary amount and expected forecast amount
- Flags for missing next steps, stale activity, unclear close dates, weak stakeholder coverage and category, amount or stage changes
- Flags for portfolio concentration when too much posture depends on one deal, owner or stage
- Recommended changes to forecast categories and follow-up actions, with the evidence gaps that limit confidence
Install the review-forecast skill
The skill is distributed as a folder named review-forecast with three files: agents/openai.yaml, references/request-schema.yaml and SKILL.md. Download the ZIP from the page where you found this guide, then unzip it into one of two places depending on how you want to use it.
For a personal install across all your projects, unzip it into ~/.claude/skills/. For a single project, unzip it into that project's .claude/skills/ folder. If you use the Claude apps instead of a local setup, you can upload the skill under Settings → Capabilities → Skills.
- Personal use: ~/.claude/skills/ (available across projects)
- Single project: the .claude/skills/ folder inside the project
- Claude apps: Settings → Capabilities → Skills, then upload the skill
agents/openai.yaml
references/request-schema.yaml
SKILL.mdRequirements: what data the skill needs
The skill needs a forecast truth source before it can build a rollup. The source names a CRM connection, a CRM report or export, or a pasted pipeline snapshot as acceptable inputs. The skill is written for a Salesforce-style CRM. If none of these is available, the skill will stop and ask for the smallest usable source rather than guess.
Other connections are optional. Knowledge files, meeting transcripts, email and internal messaging tools can be referenced through placeholder connectors. The skill uses them only for a few high-value deals where that evidence would change the call. The answers-ask-user-input skill and the ask_user_input tool are preferred for intake questions.
- Required: a CRM connection, a CRM report or export, or a pasted pipeline snapshot
- Optional: knowledge files, meeting transcripts, email and internal messaging via placeholder connectors
- Preferred for intake questions: the answers-ask-user-input skill and the ask_user_input tool
- Runs scripts: no. Uses network: yes
Trigger the skill with natural requests
You do not need special syntax. Describe the review you want in plain language, and include the scope and the source if you can. The skill is triggered by requests about forecast posture, forecast accuracy, commit, upside, forecast-call preparation, or a pipeline or deal-risk review.
If your request is vague, such as review the forecast, the skill will not assume which book or source to use. It will offer explicit choices: your current forecast-period owner book from live CRM, a named team, owner, report or pipeline view, or a pasted or exported forecast snapshot or deal set.
- Review my commit and upside for this quarter from the pipeline export I pasted
- Prepare me for the forecast call: rollup by category and stage for the EMEA team report
- Which deals in my book are at risk before my manager review on Friday?
- Check the deal-risk posture for these five opportunities: [paste names]
- Show forecast accuracy for my book using this snapshot compared with last week's
What a typical run looks like
First the skill confirms scope, period, forecast posture standard, category convention and amount basis. If any of these is only inferred, it presents a sourced default and asks you to confirm it. Then it builds the rollup from the canonical CRM report or your export, stating the exact filter, period and data freshness.
It inspects detail rows only for the highest-value, highest-risk, stale or concentration-driving deals. For the top one to three uncertain deals it adds light evidence from meetings, email or messages where that changes the call. The output ends with the risk posture, recommendation changes, evidence gaps and follow-up actions.
- Confirm scope, truth source, posture standard, category convention and amount basis
- Build the rollup by forecast category and optionally stage
- Inspect detail rows for the deals that drive the call
- Add light evidence for a few deals only where it changes the recommendation
- Deliver the posture review with evidence gaps and follow-up actions
Safety notes and limits
The initial review is read-only. The skill does not change forecast categories, close dates, CRM records, tasks or messages. Follow-up actions are suggestions, and you still have to carry them out yourself. The source says the skill uses the network, so it connects to your CRM and any optional tools you have linked. It does not run scripts.
The skill also will not treat supporting notes, messages or transcripts as a replacement for missing opportunity data in the CRM. Field history is not treated as true forecast movement unless a real prior snapshot exists.
- Read-only initial review: no CRM, close-date, task or message changes
- Uses network: yes. Runs scripts: no
- Supporting evidence cannot override CRM-owned amount, stage, category, owner or close date
- Comparisons need a real prior snapshot; field history alone does not show forecast movement
What people use it for
- Preparing a weekly forecast call with a commit and upside rollup for a team report
- A seller checking the risk on their own book before a manager review
- RevOps staff comparing a pasted pipeline snapshot against a prior snapshot for forecast accuracy
- Reviewing a named set of high-value deals for stale activity, missing next steps and unclear close dates
Frequently asked questions
How do I install the review-forecast skill?
Download the ZIP from the page where you found this guide and unzip the review-forecast folder into ~/.claude/skills/ for personal use or into a project's .claude/skills/ folder. In the Claude apps, you can instead upload it under Settings → Capabilities → Skills.
What do I need to use the review-forecast skill?
You need a forecast truth source: a CRM connection, a CRM report or export, or a pasted pipeline snapshot. The skill is written for a Salesforce-style CRM. Knowledge files, meeting transcripts, email and messaging are optional.
Is the review-forecast skill free and what is its licence?
The source lists the licence as MIT. The source does not state any price, so check the page you downloaded it from for any other terms.
Does the review-forecast skill change my CRM records?
No. The initial review is read-only and does not change forecast categories, close dates, CRM records, tasks or messages. Follow-up actions still require you to act.
Why does the skill ask me to confirm scope before it starts?
It refuses to assume an owner book, team or truth source, because a forecast built on guessed scope is unreliable. Once you confirm the scope and data source, it builds the rollup.
What should I do if the skill says it cannot build a forecast?
That usually means no usable forecast truth source is available. Provide a CRM report or export, or paste a pipeline snapshot with the opportunity fields, then ask again.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use review-forecast if you are a sales manager, RevOps person or seller who has CRM data or an export and wants a structured, evidence-aware forecast readout before a call. Pick something else if you have no pipeline data, expect the skill to update your CRM automatically, or need a tool built for a CRM other than Salesforce-style systems.
Download the skill