Opencode DelegateSAFE

Delegate a coding task to a separate coding agent CLI, review the diff, land the commit yourself — one per implementer.

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: opencode-delegate
description: >-
  Delegate a coding task to the OpenCode CLI as a background implementer, then review its diff and
  land it yourself. Use this whenever the user wants to hand implementation work to OpenCode — phrasings
  like "have OpenCode do X", "delegate this to OpenCode", "run it through OpenCode", or "use OpenCode to
  implement/fix/refactor" — or wants to run a queue of coding tasks through OpenCode while staying the
  reviewer. Prefer it when the user will review the diff and commit it themselves. DO NOT USE for tasks
  small enough to do inline, or when the user wants the code written directly without delegating.
license: MIT
compatibility: Requires the `opencode` CLI installed and authenticated, Node 18+, and git. The orchestrating agent must be able to run shell commands and read files. Shell examples assume bash/zsh (macOS/Linux, or Git Bash/WSL on Windows).
metadata:
  version: 0.5.0
---

# OpenCode Delegate

You are the **orchestrator**. This skill lets you hand a bounded coding task to a separate
**implementer** — the OpenCode CLI — then review what it produced and land it yourself. You write
the brief and own the judgment; OpenCode does the typing in its own session; you verify and commit.

Nothing here is specific to one orchestrating agent. The loop needs only the ability to run a shell
command and read a file, so any agent with those two capabilities — Claude Code, OpenCode driving a
sibling session, or a comparable one — can drive it. (It is designed for and run on Claude Code; treat
other orchestrators as designed-for, not yet proven.)

## When NOT to use this

- The task is small enough to just do inline — delegation overhead is not worth it.
- The `opencode` CLI is not installed or not authenticated (run `opencode auth login`).
- You want to write the code yourself, or you only need a review (use the `plan` agent via `--read-only`).

## Prerequisites (check once)

1. `opencode --version` succeeds. If not, install (`npm i -g opencode-ai`, or the native installer from
   opencode.ai) and `opencode auth login`.
2. **Confirm which `opencode` is on PATH.** `command -v opencode` shows the active binary and
   `opencode --version` its version. The relay records the version it ran into `result.json`, so a stale
   binary is visible after the fact.
3. A model provider is authenticated — `opencode auth list` shows at least one credential.
4. You are in (or will point `--cd` at) the target git repository.

## Choose the implementer model

OpenCode has **no safe default** — a bare `opencode run` errors — so a fresh run needs a model via
`--model` or a fleet `--lane` that sets one (a resumed run inherits its session's model). Naming the
model is the one decision a single-model backend like codex-delegate never had, and it has two owners:

- **The human owns which models are allowed.** `opencode models` lists hundreds of entries, most billed
  per token (OpenRouter and the like); only the human knows which are their flat-rate subscriptions, and
  the CLI can't tell them apart. So the usable set is theirs — ideally stated once in the repo's
  `AGENTS.md` or their `CLAUDE.md` (e.g. "delegate mechanical work to `opencode-go/...`, hard logic to
  `...`").
- **You, the orchestrator, pick per task — from that set.** Match the model to the brief: a cheap, fast
  model for a mechanical sweep (rename, migration, removal); a strong one for a subtle bug or a
  money/security path.
- **If no usable set is stated, ask — don't guess.** Guessing from the catalog risks a metered model and
  a surprise bill. Name the constraint to the human and let them choose.

More depth: [references/writing-the-brief.md](references/writing-the-brief.md).

## The loop

Run these five steps per task. Steps 1, 4, and 5 are your judgment; 2 and 3 are mechanical.

### 1. Write the brief

OpenCode sees **only** the text you send plus what it can read from the working tree — no chat history,
no shared context. Everything the task needs goes in the brief: the goal, the current state, what to
change, what to leave untouched, the project's **actual** gate commands (discover them from the repo's
AGENTS.md/CLAUDE.md/Makefile — do not assume), and a report contract. Tell OpenCode it will **not**
commit (you will). Keep one task per brief. Full guidance and a template:
[references/writing-the-brief.md](references/writing-the-brief.md).

### 2. Dispatch

For **Codex orchestration of background work**, install and configure the separate
**codex-background** support utility first. If its MCP tools or registry are missing, stop and
report the setup requirement; do not silently fall back to model-driven polling. Then invoke its
registered `delegate_run` MCP tool once in place of the shell example below.
Follow that skill for structured common arguments, unchanged provider flags and reattachment by
runId. Keep the request pending until completion, then continue this skill's review and landing
steps. Other orchestrators keep the shell workflow below.

Send the brief to OpenCode with the bundled helper. It wraps `opencode run`, captures the run, and
writes a structured `result.json` — so your only job is "run a command, read a file." (`<skill-dir>`
below is this skill's installed directory — the folder containing this `SKILL.md`. Claude Code prints
it as "Base directory for this skill" when the skill loads; on other orchestrators use that same
directory — if unsure where it landed, run `find ~ -name relay.mjs -path '*opencode-delegate*'` and
substitute the directory above it.)

```bash
node "<skill-dir>/scripts/relay.mjs" --brief brief.txt --model <provider/model> --cd /path/to/repo
# --model (or a --lane that sets model) is required on a fresh run
# fleet lane from delegate-setup:           add --lane <name>  (dials apply; flags still win)
# read-only (review/diagnosis, no edits):   add --read-only   (uses the plan agent)
# continue the previous OpenCode session:   add --resume-last  (delta brief only;

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

What the source does

Filesystem

declared (1 observation(s))

Network

none-observed

Shell

declared (2 observation(s))

Dependencies

pinned

Secrets in source

none-found

Findings (1)

const script = join(dirname(fileURLToPath(import.meta.url)), "../../delegate-setup/scripts/lane.mjs");

Gates applied: no_behavioural_pass.

[![ModelsAtlas audit: Opencode Delegate](https://modelsatlas.com/skills/amelnagdy/opencode-delegate/badge.svg)](https://modelsatlas.com/skills/amelnagdy/opencode-delegate/)

Not your listing, or something out of date? Claim or update your listing.