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.
npx agentmods add commands/simonrowland/goal-flight/decompose-plangit 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/decompose-plan)<a href="https://agentmods.dev/commands/simonrowland/goal-flight/decompose-plan"><img src="https://agentmods.dev/badge/commands/simonrowland/goal-flight/decompose-plan.svg" alt="Measured on agentmods" 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 | $0.00022 | $0.04961 |
| Opus 5 | $0.00011 | $0.02481 |
| Sonnet 5 | $0.00004 | $0.00992 |
| Haiku 4.5 | $0.00002 | $0.00496 |
Grade C, and why
decompose-plan scanned grade C with 1 finding 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 today.
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.
Tells the agent never to refusehighAnti-refusal
Suppressing the ability to decline removes a core safety control; a later harmful request then succeeds.
**Why this is right**: see `protocols/premises.md`. The whole point of premise-checking is to surface invisible assumptions for validation rather than acting on them silently. A rigid "refuse if goal isn't pinned in a sp How it starts
The opening of the file, as written. The whole thing — 279 lines — stays where its author put it; the contents beside it link to each section on GitHub.
decompose-plan []
Read:
protocols/premises.mdprotocols/dispatch-routing.mdprotocols/worker-context-package.mdprotocols/scout.md
Break a plan into numbered /goal chunks. The plan source can come from:
- A file path argument —
/goal-flight decompose-plan docs-private/refactor-plan.md - An in-context conversation — no arg; the plan was discussed in this session (the user reviewed the plan with you and it's in the chat log)
- A file Read'd earlier in this session — no arg; you find the most relevant from session context
If the source is ambiguous, ask the user which plan they want decomposed before proceeding.
Steps
0. Anchor against available context (read, don't gate)
The goal here is not to enforce a formal goal-statement gate — that would turn the goal-statement into the same kind of invisible-assumption document the skill exists to surface. The goal is to anchor the decomposition against whatever signal the user has already given about what they actually want, with explicit visibility into anything inferred.
Read whatever exists, in priority order:
- A pinned goal-statement at
<repo-root>/docs-private/goal-<topic>-*.md(or legacy<topic>-goal-statement-*.md). Read as the canonical anchor when present. - The plan source itself — file argument, in-session conversation, or referenced architecture doc (handled in step 1).
- Conversation context — anything the user told you about intent / success criteria / non-goals in this session.
Handle the goal-statement's Status: field as information, not as a refusal trigger:
Status: CONCRETEor absent (= concrete-by-default): proceed normally; keep content in mind for steps 4.5 and 6.Status: DRAFT — <reason>: proceed, but surface a high-priority inline-office-hours backlog item ("Goal-statement is DRAFT (<reason>). The decomposition will proceed on the plan source + conversation; do you want to interrogate the goal first via/office-hours, or sharpen the file directly, or just steer mid-run as decomposition surfaces the implicit-goal questions?"). Don't bail.- Goal-statement absent entirely: proceed on the plan source / conversation; surface a backlog item ("No pinned goal-statement found. The decomposition will use the plan source and our conversation as the anchor; if you want a more durable anchor across compactions, recommend
/goal-flight init <topic>later — for now, proceeding."). Don't bail.
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.
- today Changed · +2 lines · +14 tokens per session fbe141f83ec3
- 5d ago First seen · 277 lines · 8 tokens per session scan C ffdf338a8ee5
decompose-plan is a command published in the GitHub repository simonrowland/goal-flight (20 stars, last pushed today), licensed MIT. It adds 22 tokens to every session and 4,961 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it C with 1 finding (tells the agent never to refuse). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
review-context7-updates
Please read the @README.md and @docs/tools/package-documentation.md to understand the high level context of my repo.
mine
Quickly capture relevant project context into Memento memory.
help
Show comprehensive Memento help — available MCP tools and usage patterns.
init
Set up Memento — configure MCP server and verify everything works.
search
Search your memories using Memento hybrid retrieval.
status
Show Memento server status and current configuration.