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.
git clone --depth 1 https://github.com/simonrowland/goal-flightWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/simonrowland/goal-flight/init)<a href="https://agentmods.dev/commands/simonrowland/goal-flight/init"><img src="https://agentmods.dev/badge/commands/simonrowland/goal-flight/init/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/simonrowland/goal-flight/init"><img src="https://agentmods.dev/badge/commands/simonrowland/goal-flight/init.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00024 | $0.03234 |
| Opus 5.5 | $0.00010 | $0.01294 |
| Sonnet 5.5 | $0.00005 | $0.00647 |
| Haiku 4.5 | $0.00002 | $0.00323 |
Grade A, and why
init scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 27d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 285 lines — stays where its author put it; the contents beside it link to each section on GitHub.
init
Initialize a project for goal-flight with compact, procedural discovery.
Read protocols/session-preflight.md, protocols/tool-readiness.md,
protocols/project-state-layout.md, protocols/task-lifecycle.md,
protocols/progress-dashboard.md, protocols/premises.md,
protocols/state-handoff.md, protocols/worker-context-package.md,
protocols/scout.md, and protocols/chunk-review.md (review tooling).
Steps
- Confirm host setup/bootstrap happened. If this is a fresh host install, run
setup from your Goal Flight clone (default
~/.goal-flight) before project init:
./setup.sh --agent codex
./setup.sh --apply --yes --agent codex
./setup.sh --cursor
./setup.sh --apply --yes --cursor
# Optional Cursor project-local install:
./setup.sh --apply --yes --cursor-project "$PWD"
After source SKILL.md, commands/, protocols/, templates/, or adapters/
changes, copied host installs need a resync from the source repo with
./install.sh <host> or the matching setup path unless the host skill path is a
symlink. Doctor JSON reports installed_skill_drift, and text mode prints
installed_skill_md_hash WARNs.
For Codex, setup registers the Goal Flight package for the Desktop orchestrator
surface, cleans any duplicate legacy personal Codex skill when plugin
registration succeeds, checks the CLI worker surface, and registers
context-mode MCP when needed. codex exec remains the worker path. For Cursor,
setup installs global agent instructions, a personal skill, rules, and Cursor
MCP config for context-mode. Project-local setup writes the wrapper and, when
context-mode is selected, project-specific .cursor/mcp.json under the target
repository.
Default add-ons (same tier as setup prompts):
- context-mode — large-output offload
- gstack — default independent reviewer for both chunk-level pre-commit
review and milestone review. When host gstack skills are missing, init offers:
minimal subset (default:
/review,/office-hours,/plan-eng-review, plus downloaded community skillsgrill-mefromudecode/plateandthermo-nuclear-code-quality-reviewfromcursor/plugins), full pack, or skip. Full-pack delegation is used only for hosts the upstream gstack setup supports; unsupported hosts fall back to the minimal copy path with a teaching message. The minimal path exposes only the Goal Flight subset from an existing local gstack checkout/cache and downloads the two public community skills only after the same consent prompt; download failures warn and degrade gracefully. Download sources are pinned HTTPS URLs; the test-only source override (GOALFLIGHT_GSTACK_EXTERNAL_SOURCE_*) is ignored unlessGOALFLIGHT_ALLOW_EXTERNAL_SOURCE_OVERRIDE=1is also set, with a visible ignored-override warning otherwise. Local fallback prompts live inprompts/gstack-claude-review.mdandprompts/gstack-codex-challenge.mdwhen gstack is not installed. - autoreview — complementary diff-local pre-commit pass via
scripts/autoreview.sh. Runs in parallel with gstack at chunk level when the orchestrator chooses; does not replace gstack as the default review path. Vendored atautoreview/; doctor WARNs ifautoreview/scripts/autorevieworscripts/autoreview.shis missing (override helper withAUTOREVIEW_HELPER).
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.
- 27d ago Changed · +5 lines c9769f27f3ce
- 29d ago Changed · +1 lines 564dd5f9495f
- 1mo ago Changed · +4 lines · +16 tokens per session 4c945de325cb
- 1mo ago First seen · 275 lines · 8 tokens per session scan A 4c81eab7c5b9
init is a command published in the GitHub repository simonrowland/goal-flight (20 stars, last pushed 22d ago), licensed MIT. It adds 24 tokens to every session and 3,234 once invoked, about $0.0001 per session on Opus 5.5. 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.
Other commands, from other repositories
pm-gate
Explicit gate approval for great-pm. Under gate-policy: explicit (great-pm's default), agents NEVER auto-close gates — only this command does. Subcommands: approve | reject | list | show .
pm-help
List all great-pm commands and what each does, grouped by stage / purpose.
pm-agent-retire
Mark a great-pm agent as deprecated. Adds a deprecation banner to the agent file, files a Beads issue to remove dependent commands/workflows, and logs the rationale. NEVER deletes the file (history preserved).
pm-audit
Run the great-pm PM-health audit on a product, an initiative, or great-pm itself. 15 dimensions, severity-rated findings, Top-5 + Quick-Wins + Things-Look-Bad-But-Fine + Open Questions. Files findings as Beads tasks unless --read-only.
pm-discover
Kick off the Discover stage for an existing initiative. Spawns user-researcher + feedback-synthesizer + market-analyst IN PARALLEL. Outputs three drafts in .great-pm/drafts/.
pm-experiment
Design an A/B test, holdout, multivariate, or switchback experiment. Spawns experiment-designer. Output is gate-quality: pre-registered hypothesis, MDE, guardrails, stopping rules.