Skip to content
Agentmart
Submit a tool

enterprise-build

Use when a locked enterprise contract exists and implementation must proceed through strict TDD, contract traceability, and staged review handoffs

Skill by anotherben

IndexedMIT
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/anotherben-enterprise-build

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Use when a locked enterprise contract exists and implementation must proceed through strict TDD, contract traceability, and staged review handoffs

Use cases: documents, testing

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/enterprise-build

What's inside

Skill name
enterprise-build
Description
Use when a locked enterprise contract exists and implementation must proceed through strict TDD, contract traceability, and staged review handoffs
License
MIT
Instructions
51 lines in SKILL.md · view on GitHub

Files (3)

  • SKILL.md
  • evals.json · evals/
  • build-status-tracking.md · references/

SKILL.md (first lines)

# Enterprise Build

Contract-driven implementation for portable enterprise work.

## Required Background

- `test-driven-development`
- `subagent-driven-development` when the plan has safe isolated tasks

## Entry Gate

Before editing source files, confirm:

- the current agent session recorded the plan
- the current agent session recorded the contract
- the recorded contract status is `LOCKED`

If any of those are missing, stop and go back upstream.

## Required Workflow

1. Read the locked contract, the implementation plan, and the current repo profile.
2. Run the agent-bound build gate in [agent-stage-gates.md](../enterprise/references/agent-stage-gates.md) before editing source files.
3. Implement one postcondition at a time with TDD.
4. Record progress in the build status model from [build-status-tracking.md](references/build-status-tracking.md).
5. Record build-stage handoff data in the current agent session before review.
6. Do not add scope that is not in the contract.
7. Run or stage the repo gate commands named by the contract as soon as their prerequisites exist, not only after PR creation.
8. Hand off to `enterprise-review` after each meaningful task boundary or after the whole build, depending on the path size.

For schema/query/data-sensitive work, the first relevant RED test must be live DB or real integration proof against migrated Postgres. Mocked DB tests may supplement behavior coverage but do not satisfy schema/query proof.

For UI, PDF upload, file upload, preview/download, modal, navigation, or rendered-output work, build must leave a headless browser command ready for review/verify.

For repos with no-new-mock or DB ownership gates, build must keep the diff compliant while coding. Do not introduce new inline DB mocks for schema/query proof, and do not route writes through mixed-owner direct-write files unless the locked contract explicitly owns that seam.

If the locked contract already exists and the prompt asks for the next step, classify the transition as `STAGE_ONLY` build entry rather than re-triaging the whole program.

If the prompt explicitly states that the committed repo profile and repo-local overlay are already current, do not reopen discover work by default.

## Rules

- One `RED` -> one `GREEN`
- The test must fail for the right reason before code is added
- Contract amendments only happen through the forge recycle loop
- If a planned task is independent, isolated-task execution is preferred over one large mixed diff
- For bounded work, keep build-status notes terse and leave enough runway for review and verification in the same session
- CI-driven code pivots invalidate previous review/forge coverage for the affected files. Re-run review/forge on the current diff after the pivot.
- Avoid brittle assertions on volatile implementation constants. Idempotence, ranking, tuning, heuristic, or generated-score tests should prove invariant behavior, state transitions, and no-ratchet stability unless the exact number is a locked business rule.

From anotherben/agentic-governance, MIT.

More from anotherben/agentic-governance

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
May 4, 2026 (5 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from anotherben

Cite this page

enterprise-build skill. Agent Mart. https://agentmart.dev/skills/anotherben-enterprise-build (last updated 2026-10-24)