execute

A command for running queued Goal Flight tasks with worker processes, optionally in parallel.

In plain words
What is it for?
Use it to execute planned goal chunks, request a design checkpoint before selected tasks, and run independent work in separate worktrees when parallel mode is enabled.
Why use it?
It automates dispatching the next pieces of work, but may spend money and change a working copy of the code.

Command

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/simonrowland/goal-flight/execute
Clone the repo
git clone --depth 1 https://github.com/simonrowland/goal-flight
Per session 9 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 4,035 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00009 $0.04035
Opus 5 $0.00005 $0.02018
Sonnet 5 $0.00002 $0.00807
Haiku 4.5 $0.00001 $0.00404

Measured yesterday against content hash c3578ad7d236, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

execute scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

commands/execute.md · 346 lines

How it starts

The opening of the file, as written. The whole thing — 346 lines — stays where its author put it; the contents beside it link to each section on GitHub.

execute [--parallel ]

Execute the next goal-queue chunks with procedural workers and compact status.

DISPATCHES WORKERS. Spawns worker processes, leases capacity (costs money), and may mutate a worktree. See "Command danger classification" in SKILL.md. The standing com.goalflight.drain daemon (every ~60s, always-on) launches anything queued — draining is automatic, not a manual step.

Read:

  • protocols/session-preflight.md
  • protocols/dispatch-routing.md
  • protocols/worker-markers.md
  • protocols/worker-context-package.md
  • protocols/scout.md
  • protocols/worker-contract.md
  • protocols/state-handoff.md
  • protocols/user-status-cadence.md
  • protocols/chunk-review.md
  • protocols/milestone-review.md
  • protocols/worktrees-parallel.md only for --parallel

Two things that decide whether a dispatch is worth its tokens

Ask for a plan before the build on CP/H-class chunks — by tagging the brief. Tag it [design-checkpoint]. The worker's first deliverable is then the design — interfaces, invariants, acceptance checks, intended test names — written to the lane's research directory, followed by !USER-NEED: design ready for approval at <path>; session <resume-id>; log <path>, and it stops. Resume that session with your corrections; do not re-dispatch — re-dispatching re-pays the orientation the checkpoint exists to save. A wrong design then costs one steer instead of a whole build plus a rewrite.

This is opt-in per brief, and deliberately so. It is NOT on by default and must not be: !USER-NEED: ends the run, so a worker that checkpoints on its own initiative during an unattended execute stalls the queue until a human appears — and because the background wait correctly wakes you on that terminal, the stall looks like normal progress. Only the codex goal-mode template carries the rule today; the dispatch-wrapper path does not, so tag the brief rather than assuming the worker was told.

Collect approvals in a batch, not one at a time. A claimed controller keeps its ownership listener armed and does not enumerate dispatch ids; after each wake, use owned status to collect compatible terminal checkpoints. For an unclaimed scripted fixed-set join only, --wait takes several ids and returns when all are terminal, and a design checkpoint is terminal:

Read the full file on GitHub · 346 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 346 lines · 9 tokens per session scan A c3578ad7d236

Subscribe to this mod's changes

execute is a command published in the GitHub repository simonrowland/goal-flight (20 stars, last pushed 3d ago), licensed MIT. It adds 9 tokens to every session and 4,035 once invoked, about $0.0000 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.