copilot-delegate

Copilot Delegate

You are the orchestrator. Delegate a bounded coding task to a separate implementer — the GitHub Copilot CLI — then review what it produced and land it yourself. You write the brief and own the judgment; the implementer makes changes in its own session in a clean working tree; you verify and commit.

The loop needs only a shell command and file access, so any comparable orchestrator can drive it.

When NOT to use this

• The task is small enough to do inline; delegation overhead is not worth it.
• The copilot CLI is not installed or authenticated.
• You need a hard sandbox. Copilot exposes sandbox controls, but they are upstream-experimental

(MXC-based, controlled via the /sandbox command and settings, disabled by default) — this relay does not configure them. --read-only only disables edit tools (--mode plan); shell commands still run. If project files must not change at all, dispatch against a clean or isolated worktree.

Prerequisites (check once)

1. Install copilot (npm install -g @github/copilot; the CLI requires Node 22+, the relay

itself runs on Node 18+ — the relay probes copilot version).

1. Authenticate: run copilot login (interactive web/device flow), or set

COPILOT_GITHUB_TOKEN / GH_TOKEN / GITHUB_TOKEN in the environment.

1. Confirm copilot version succeeds.
2. Work in, or point --cd at, the target git repository.

Choose the model (optional)

Copilot picks a default model (auto). To choose another, pass --model <name>. The relay accepts letters, digits, and . _ : / - only (the value reaches a shell on Windows).

Choose the effort (optional)

Copilot supports a reasoning effort dial: --effort <level> with values low, medium, high, xhigh, or max. The relay rejects any other value before dispatch.

The loop

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

1. Write a brief

Copilot sees only the text you send. It cannot read your conversation: the brief must stand alone with the goal, current state, what to change, what to leave untouched, the project's real gates, and a report contract. Keep each brief to a single task. Write it to a file and pass it as the relay's --brief. See references/writing-the-brief.md.

2. Dispatch

Use the bundled relay. It runs copilot -p with --output-format json --no-color --stream off, captures the JSONL event stream, and writes result.json.

node "<skill-dir>/scripts/relay.mjs" --brief brief.txt --cd /path/to/repo
# choose a model:                    add --model <name>
# set reasoning effort:              add --effort <level>
# read-only planning pass:           add --read-only  (forces --mode plan)
# full tool autonomy:                add --allow-all-tools
# hard time limit (watchdog):        add --timeout 2h   (the 30m default suits brief runs)
# resume a session:                  add --session <id>  or  --resume-last
# see all options:                   node .../relay.mjs --help

The child's cwd pins the workspace. The relay writes artifacts under the system temp dir by default and never commits. See references/dispatch-and-poll.md.

3. Wait for completion

The relay blocks until copilot finishes. Run it with the orchestrator's background-command facility, or background it in the shell and poll for result.json. A pre-run usage error exits 2 and writes no result; a missing copilot exits 127 and writes status: "copilot_unavailable".

Completion means the process exited and result.json exists — trust process state and the working tree, not the progress display. Copilot's final assistant message is the finalMessage field of result.json.

4. Review — do not trust the self-report

• Re-run the project's gates yourself.
• Read the diff against the brief, starting with touchedFiles.
• Run relevant guard skills if installed.

See references/review-and-land.md.

5. Land it

If the work is good, commit it. The relay never commits — the diff and result.json are the record; run git status and git diff first to confirm exactly what changed. If the group has a PR flow, make the commit and push a branch; let human review happen. If the diff is wrong or incomplete, re-dispatch a corrected brief in a fresh run and review again.

Autonomy and permissions

Without --allow-all-tools, copilot auto-denies tool calls in headless mode: the process exits 0 but the relay detects the denial events and reports status: "failed" with the CLI's own error message and a hint to pass --allow-all-tools. This is the honest default — the orchestrator sees the failure rather than a silent no-op.

--allow-all-tools explicitly grants full tool autonomy. --read-only selects --mode plan, which disables edit tools so project files can't be changed by direct edits; it works without --allow-all-tools. Shell commands still run in plan mode, so it guards against edits, not against everything. The two flags are mutually exclusive.

Copilot also exposes sandbox controls, but they are upstream-experimental (MXC-based, controlled via the /sandbox command and settings, disabled by default). This relay does not configure them.

Authorization model

Delegation is something the human opts into. Once briefed, copilot works as a tool you approved use of. The boundary is: do not accept conclusions from the self-report; verify everything on disk. For anything touching credentials, production data, or irreversible operations, stop and ask the human first instead of encoding it in a brief.

References

• references/writing-the-brief.md — structure, scope, gates,

brief delivery.

• references/dispatch-and-poll.md — flags, artifacts,

result.json, and failure recovery.

• references/review-and-land.md — what to verify before calling

the diff done, at the end of a run.

• references/multi-task-queues.md — sequential queues,

constraint carry-forward, progress tracking, and the final coherence pass.