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/madarauchiha-314/the-loop/askgit clone --depth 1 https://github.com/MadaraUchiha-314/the-loopWrote 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/madarauchiha-314/the-loop/ask)<a href="https://agentmods.dev/commands/madarauchiha-314/the-loop/ask"><img src="https://agentmods.dev/badge/commands/madarauchiha-314/the-loop/ask.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 | $0.00000 | $0.00605 |
| Opus 5 | $0.00000 | $0.00302 |
| Sonnet 5 | $0.00000 | $0.00121 |
| Haiku 4.5 | $0.00000 | $0.00060 |
Grade A, and why
ask 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 4d 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 — 53 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ask
Ask a human a question on the work item — the way a spawned agent escalates.
the-loop ask --work-item github:OWNER/REPO#N --question "Which auth mode?"
the-loop ask --work-item github:OWNER/REPO#N --question-file question.md
some-tool | the-loop ask --work-item github:OWNER/REPO#N --question-file -
What it does
One verb, three effects — in this order:
- Posts the question as a comment on the work item through your own
gh(the issues endpoint serves PR conversations too), with the visible attribution line and the loop-prevention marker appended centrally. That stamping is the point (issue-208): before this verb, every agent was trusted to remember the marker on every question, and one lapse turned its own comment into an event that resumed its own session. - Records the wait as a
session.awaiting_inputevent — question text, the comment's URL, who asked. This is what makes the waiting session observable:GET /api/v1/attentionreports the work item asawaiting-input, and the dashboard's question card lights up. - Leaves the answering to the reply route: an operator answers either on the
ticket (the poller forwards it, as ever) or straight into the session's pane
via
POST /api/v1/sessions/reply, which emits thesession.reply_sentthat closes the wait.
If gh fails, the exit code is 1 and the failure is printed — but the wait is
still recorded (comment_posted: false, level warning): the agent is
waiting whether or not GitHub was reachable, and the reply route can still carry
the answer.
Flags
| Flag | Default | Meaning |
|---|---|---|
--work-item |
required | The ref the question goes to. |
--question |
— | The question text (markdown). |
--question-file |
— | Read the question from a file; - reads stdin. Exactly one of the two. |
Notes
- Runs in-process, not through the control-plane service — deliberately, and unlike the other session verbs: this is the escalation path, the one verb that must keep working when no service is running (or in a cloud session that has no daemon at all).
- An empty question or malformed ref exits 2 and posts/records nothing.
- The stamp is idempotent: a question that already carries the marker is not double-stamped.
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.
- 4d ago First seen · 53 lines · 0 tokens per session scan A dc540a67c534
ask is a command published in the GitHub repository MadaraUchiha-314/the-loop (4 stars, last pushed today), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 605 tokens. 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
gaai-status
Show backlog, memory state & delivery readiness.
gaai-daemon
Launch or inspect the GAAI Delivery Daemon.
gaai-deliver
Launch or inspect the GAAI Delivery Daemon.
help
Show help for Ralph Specum plugin commands and workflow.
new
Create new spec and start research phase.
discover
Run a full product discovery cycle — from outcome definition through opportunity mapping, prioritisation, and experiment design. Use when the team isn't sure what to build next, or before writing a PRD for a complex feature space.