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/Habib0x0/spec-driven-pluginWrote 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/habib0x0/spec-driven-plugin/research)<a href="https://agentmods.dev/commands/habib0x0/spec-driven-plugin/research"><img src="https://agentmods.dev/badge/commands/habib0x0/spec-driven-plugin/research/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/habib0x0/spec-driven-plugin/research"><img src="https://agentmods.dev/badge/commands/habib0x0/spec-driven-plugin/research.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 | $0.00029 | $0.01594 |
| Opus 5.5 | $0.00012 | $0.00638 |
| Sonnet 5.5 | $0.00006 | $0.00319 |
| Haiku 4.5 | $0.00003 | $0.00159 |
Grade A, and why
research 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 2mo 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 — 114 lines — stays where its author put it; the contents beside it link to each section on GitHub.
$ARGUMENTS
Research this thoroughly before any planning or implementation begins.
How to research
Step 1: Clarify before you research (MANDATORY — never skip)
Before reading a single file or launching any agent, use AskUserQuestion. Read the input and identify every place where you have 2+ plausible interpretations — scope, intent, constraints, approach, priority. Ask about those specifically.
How to ask: Present choices tailored to the actual input, not generic categories. The options should come directly from the ambiguities in what was asked. If you see three plausible ways to interpret what the user wants, list those three things and ask which is closest. Don't ask what you can already infer. Do ask anything that would materially change what you research or recommend.
Good trigger conditions for asking:
- The input describes a symptom but not a root cause — ask what they think the cause is, with options
- The input proposes a solution — ask if the solution is required or just a starting hypothesis
- The scope is fuzzy — ask whether they want a targeted fix or a broader rethink, with examples of each
- Multiple approaches exist with real tradeoffs — ask which tradeoffs matter most to them
- The change could affect related systems — ask whether those are in scope
- Any constraint (time, backwards-compat, file/dependency, team conventions) is unstated — ask
Keep questions short. Use choices and options, not open prompts. "Which of these is closer?" beats "Can you describe your constraints?". An "other/none of these" escape hatch is always fine to include.
Ask as many questions as the ambiguity warrants — but batch them into a single AskUserQuestion call so the user responds once.
Do not launch any agents until you have the answers.
Step 2: Parse intent
With the answers in hand, read critically:
- What is the core problem — distinct from the proposed solution?
- Does any answer change the scope or approach from what was originally described?
- Are there remaining ambiguities? If yes, use AskUserQuestion again — don't bank on assumptions.
- Frame 2-4 specific research questions around the problem.
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.
- 2mo ago First seen · 114 lines · 29 tokens per session scan A ce6079e2ddb2
research is a command published in the GitHub repository Habib0x0/spec-driven-plugin (10 stars, last pushed 2d ago), licensed MIT. It adds 29 tokens to every session and 1,594 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
brief
Brief a PR, MR or the current branch with Punchcard — the reviewer's orientation before the review, never a review.
pr
Review a GitHub PR or GitLab MR with Punchcard and post the review into it.
re-review
Round 2+ review — confirm revisions landed, catch regressions, do not re-litigate.
review-spec
Unbiased review of another session's spec or design doc.
scaffold-component
Scaffold a single named component from the architecture blueprint with production-depth code.
design-system
Generate a complete design system from product context and wireframes — no blueprint required.