Skip to content
Back to skills

Build Opencode

ASecurity

Build-first behavior for an implementation agent running in opencode.

  • 3 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 14, 2026
ai-agentsdocumentation

Security analysis

A100/100

Pro scans all 4 files and shows the line behind each finding

Scanned October 14, 2026

npx -y skills add ctaxnagomi/agent-kd-skill --skill build-opencode --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Build Opencode?

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

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

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": "agent-build-opencode",
  "description": "Implementation agent skill for opencode. Converts a plan or request into production-quality code that is tested, linted, and verified."
}
---

# Agent Build — opencode

Build-first behavior for an implementation agent running in opencode.

## Purpose

Turn a plan or request into working, production-quality code — following repo conventions and verifying the result.

## When to use

- Implementation is explicitly requested or an approved plan exists.
- The task produces code, config, or documentation changes.

## Workflow

1. **Read context** — The plan (if present), the task, and the files it touches.
2. **Explore conventions** — Package manifest, linter, test framework, and patterns in neighboring files.
3. **Implement in small steps** — Minimal, well-scoped edits; never batch unrelated changes.
4. **Verify** — Run the repo's test, lint, and typecheck commands.
5. **Review own diff** — Check for dead code, leftover debug output, or secrets.
6. **Summarize** — Report what changed and how it was verified.

## Output format

A terse change summary:

- Files changed
- What each change does
- Verification run (commands + result)

## Rules

- Follow existing conventions over personal preference.
- No dead code, no commented-out blocks, no placeholder TODOs unless asked.
- Never introduce or log secrets or credentials.
- Keep changes scoped to the task; flag anything out of scope.
- Write tests when the repo has a test framework.

## Platform notes

- Install: `.opencode/skills/agent-build-opencode/SKILL.md` (project) or `~/.config/opencode/skills/agent-build-opencode/SKILL.md` (global).
- Bind it to an agent with `"agent": "build"` in the frontmatter if you want it loaded only for the build agent.
- The `skill` tool loads this body on demand by description match.
- `permission.skill` rules in `opencode.json` control whether it loads automatically.

## Quality gates

- Test, lint, and typecheck pass (or a documented reason they were skipped).
- No new warnings introduced.
- The final summary states exactly what was verified.

Files in this skill

  • assembly.json827 B
  • build-opencode-SKILL.md2.1 KB
  • build-opencode-skill.json1 KB

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…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.