Skip to content
Back to skills

Build

ASecurity

Code generation from specifications, architecture, or requirements

  • 3 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 1, 2026
ai-agentspythongojavabashreactnodefastapiapidocumentation

Works with

  • api

Security analysis

A100/100

Scanned October 1, 2026

npx -y skills add sharmapuneet1510/awesome-prompts --skill build --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Build?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Build
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/sharmapuneet1510-build/badge)](https://www.skillsdirectory.com/skills/sharmapuneet1510-build)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: build
description: Code generation from specifications, architecture, or requirements
argument-hint: path="..." [tech_stack="..."]
disable-model-invocation: true
---

Role and rules: read ${CLAUDE_PLUGIN_ROOT}/reference/agent.md and ${CLAUDE_PLUGIN_ROOT}/reference/rules.md for the sections this function needs.

# implementer:build

**Generate production-ready code** from specifications, architecture documents, or requirements.

## Inputs

```
implementer:build path="..." [tech_stack="..."]
```

- `path` (string, required) — Path to spec, design, or requirements file
- `tech_stack` (string, optional) — Target tech (Java, Python, React, Node.js, etc.)
- `style` (string, optional) — Code style (conventional, google, airbnb)

## Outputs

```
✓ src/                        — Generated source code
✓ ARCHITECTURE.md             — Code structure documentation
✓ BUILD.md                    — Build and deployment guide
```

## Example

```bash
implementer:build path=./api-spec.yaml [tech_stack="Python, FastAPI"]
```

**Output:**
- FastAPI routes with type hints
- Pydantic models
- Docstrings
- Error handling

## Gate — Approved Spec Only (RULE 11 / 11a)

Run these checks **before** generating any code. Each failure is a refusal
with a named next step, not a warning to proceed past.

| Check | Refuse when | Next step |
|---|---|---|
| Spec chain | The chain is not approved (the gate (`engineering-skills:spec-driven-development` skill)) | `orchestrator:plan` → `architect:design` |
| ADR gate | The change alters a contract, data shape, dependency, or failure mode and no Accepted ADR cites its Parent Jira | `architect:adr` |
| Traceability | `ba:trace` reports a High finding on T-1, T-2, T-10, or T-13 | Resolve the finding |
| Scope | A task has no `REQ-n` citation | `orchestrator:plan` |

Implement **only** what the approved tasks and Accepted ADRs describe. Work
not traceable to a requirement is scope creep, and `quality:observe` will
report it as such.

Trivial work — one-line fixes, config tweaks, small doc edits — is exempt per
RULE 11.

## Workflow

0. Run the gate above. Refuse if any check fails.
1. Parse specification/design
2. Read `docs/adr/` for Accepted decisions constraining this area, and
   `docs/project-context/` for invariants and known behaviours
3. Detect or use specified tech stack
4. Generate code structure
5. Implement functions/classes — honouring every Accepted ADR; contradicting
   one requires `architect:adr` to supersede it first
6. Add error handling
7. Generate documentation
8. Create build configuration
9. Update `docs/project-context/dependencies.md` if dependencies changed

## Related Functions

- `implementer:full` — Build + test + doc in one pass
- `implementer:test` — Test generation
- `implementer:doc` — Documentation generation
- `architect:design` — System design (input)
- `architect:adr` — Required before decision-bearing changes
- `ba:trace` — Gate checks

## Related Skills

- `skills/spec_driven_development_skill.md` · `skills/adr_skill.md`
- `skills/traceability_skill.md` · `skills/project_context_skill.md`

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…