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/GOAL-T/goaltreeWrote 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/goal-t/goaltree/start)<a href="https://agentmods.dev/commands/goal-t/goaltree/start"><img src="https://agentmods.dev/badge/commands/goal-t/goaltree/start/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/goal-t/goaltree/start"><img src="https://agentmods.dev/badge/commands/goal-t/goaltree/start.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 | $0.00023 | $0.00386 |
| Opus 5.5 | $0.00009 | $0.00154 |
| Sonnet 5.5 | $0.00005 | $0.00077 |
| Haiku 4.5 | $0.00002 | $0.00039 |
Grade A, and why
start 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 1mo 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.
What it actually says
Build a GoalT goal tree for the current project.
- Figure out the absolute path of the project root (the current working directory, unless told otherwise).
- Explore the codebase before inventing any structure: read the README if one exists, the package manifest (package.json / pyproject.toml / etc.), the top-level folder layout, and -- if present -- database migration files or a schema definition (e.g. a
supabase/migrationsfolder, a Prisma schema, a Django models file). - From what you actually find, identify the project's real functional areas (e.g. authentication, payments, a specific feature area) -- don't guess generic categories that don't match this codebase.
- Call
create_treewith a root_label describing the project andproject_rootset to the absolute path from step 1. - Call
add_goalfor each functional area you identified, writing a genuine one-or-two-sentence description of what it does in this specific codebase. Where you can confidently identify the files that implement it, pass them asrelated_files(paths relative to project_root). Where relevant backend artifacts exist (database tables, edge functions, API routes), pass them asrelated_backend. - If some goals genuinely depend on or serve more than one parent area, give them multiple parents -- don't force everything into a single-parent tree if the real structure isn't like that.
- Once the tree is built, call
open_dashboardand tell the user it's ready, briefly summarizing what you found.
Do not fabricate files or backend artifacts you haven't actually seen -- it's fine, and expected, for some goals to have no related_files yet if you're not confident about the mapping. Leave those empty rather than guessing.
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.
- 1mo ago First seen · 16 lines · 23 tokens per session scan A f971d8db3705
start is a command published in the GitHub repository GOAL-T/goaltree (0 stars, last pushed 17d ago), licensed Apache-2.0. It adds 23 tokens to every session and 386 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 2026-08-31.
Other commands, from other repositories
contact-upsert
Create or update a contact, optionally assign to groups.
brainstorm
Brainstorm product ideas or experiments from PM, Designer, and Engineer perspectives — for existing or new products.
plan-okrs
Brainstorm team-level OKRs aligned with company objectives — qualitative objectives with measurable key results.
write-stories
Break a feature into backlog items — user stories, job stories, or WWA format with acceptance criteria.
specify
Create specs from project briefs with acceptance criteria and testable requirements.
prompt
System instructions for writing effective prompts. Apply when generating commands, skills, agents, or any LLM instructions.