Skip to content
Beta Multi-model support (OpenAI, Gemini, and more) launching soon.
RBA RBA Skills

/estimate

Installable Ships in estimate-sow-builder — or copy the raw command below.
Open Claude.ai

Paste into Claude Code, Claude.ai, the desktop app, or the browser extension — ⌘/Ctrl+V.

ExperimentalNo security review

/estimate — Estimate & SOW Builder

This command is the guided entry point for the full estimation-to-SOW pipeline.

Usage

/estimate                         # Interactive mode — guided prompts
/estimate build <brief>           # Run build-estimate on a brief
/estimate risk <estimate-file>    # Run risk-weight-estimate on an existing estimate
/estimate wbs <estimate-file>     # Run generate-wbs on an approved estimate
/estimate sow <estimate-file>     # Run draft-sow (optionally pass WBS file too)
/estimate co <change-description> # Run change-order-builder
/estimate review <sow-file>       # Run sow-reviewer agent on a draft SOW

Interactive mode (default)

When run without arguments, prompt the user:

Welcome to the RBA Estimate & SOW Builder.

What would you like to do?

1. Start a new estimate from a project brief
2. Apply risk weighting to an existing estimate
3. Generate a WBS from an approved estimate
4. Draft a SOW from an approved estimate
5. Build a change order for a mid-engagement scope change
6. Review a draft SOW before it goes to the client
7. Run the full pipeline (estimate → risk → WBS → SOW)

Enter a number or describe what you need:

Based on the response, invoke the appropriate skill:

Choice Skill invoked
1 build-estimate
2 risk-weight-estimate
3 generate-wbs
4 draft-sow
5 change-order-builder
6 sow-reviewer agent
7 Full pipeline in sequence

Full pipeline mode (option 7)

Run the skills in sequence with handoffs:

  1. Run build-estimate → writes estimate-[slug]-[date].md
  2. Ask user: "Review the estimate above. Shall I proceed to risk weighting? (yes / make changes first)"
  3. Run risk-weight-estimate on the estimate file → writes risk-report-[slug]-[date].md
  4. Ask user: "Risk report complete. Shall I generate the WBS? (yes / adjust estimate first)"
  5. Run generate-wbs on the estimate file → writes wbs-[slug]-[date].md
  6. Ask user: "WBS generated. Shall I draft the SOW? (yes / review WBS first)"
  7. Run draft-sow on the estimate + WBS files → writes sow-draft-[slug]-[date].md
  8. Ask user: "SOW drafted. Shall I run the SOW reviewer? (yes / review first)"
  9. Run sow-reviewer agent on the draft SOW → outputs review report

File naming convention

All output files follow the pattern: [artifact]-[project-slug]-[YYYY-MM-DD].md

Examples:

  • estimate-finserv-portal-2026-06-01.md
  • risk-report-finserv-portal-2026-06-01.md
  • wbs-finserv-portal-2026-06-01.md
  • sow-draft-finserv-portal-2026-06-01.md
  • change-order-CO-001-finserv-portal-2026-06-01.md

If no project slug can be inferred, use new-engagement.

Tips

  • Run /estimate build early in the sales cycle for a ROM. Run the full pipeline once the brief is solidified.
  • Always run risk-weight-estimate before draft-sow — the risk report often changes the total hours and the assumptions list.
  • The sow-reviewer agent is the final gate. Treat its BLOCKERs as mandatory fixes.
Technical details
Source
estimate-sow-builder
Author
Dakota Smith
Invocation
/estimate
Maturity
experimental
Security review
No security review
Owners
  • delivery
File
plugins/estimate-sow-builder/commands/estimate.md

Edit this command on GitHub — opens GitHub's web editor. No git knowledge needed.