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/Connectry-io/connectrylab-architect-cert-mcpWrote 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/connectry-io/connectrylab-architect-cert-mcp/refine)<a href="https://agentmods.dev/commands/connectry-io/connectrylab-architect-cert-mcp/refine"><img src="https://agentmods.dev/badge/commands/connectry-io/connectrylab-architect-cert-mcp/refine/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/connectry-io/connectrylab-architect-cert-mcp/refine"><img src="https://agentmods.dev/badge/commands/connectry-io/connectrylab-architect-cert-mcp/refine.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 | $0.00000 | $0.00513 |
| Opus 5 | $0.00000 | $0.00257 |
| Sonnet 5 | $0.00000 | $0.00103 |
| Haiku 4.5 | $0.00000 | $0.00051 |
Grade A, and why
refine 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 9d 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
Iterative Refinement
Collaboratively refine a specification, prompt, or code pattern through structured conversation rounds.
Topic
$ARGUMENTS
Process — The Interview Pattern
Round 1: Understand Intent
Ask 3-5 clarifying questions before generating anything. Example questions:
- What is the primary use case for this?
- Who is the audience (developers, end users, CI system)?
- Are there existing patterns in the codebase to match?
- What are the constraints (performance, compatibility, size)?
- Can you show me an example of what "good" looks like?
Round 2: Show Input/Output Examples
Present 2-3 concrete input/output pairs that demonstrate your understanding:
--- Example 1 ---
Input: createUser({ name: "Alice", email: "[email protected]" })
Output: { ok: true, data: { id: "usr_abc", name: "Alice", email: "[email protected]" } }
--- Example 2 ---
Input: createUser({ name: "", email: "invalid" })
Output: { ok: false, error: "Validation failed: name is required, email is invalid" }
--- Example 3 (edge case) ---
Input: createUser({ name: "Bob", email: "[email protected]" }) // email already exists
Output: { ok: false, error: "A user with this email already exists" }
Ask: "Do these examples match your expectations? What would you change?"
Round 3: Generate Draft
Produce a first draft based on confirmed understanding. Mark uncertain decisions with [DECISION NEEDED] tags so the user can resolve them.
Round 4: Refine
Incorporate feedback. Repeat until the user confirms the output is correct.
Guidelines
- Never assume; always ask when ambiguous
- Show concrete examples rather than abstract descriptions
- Keep each round focused on one concern
- Track decisions made in earlier rounds; do not revisit unless asked
- Converge toward a final artifact within 3-4 rounds
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.
- 9d ago First seen · 68 lines · 0 tokens per session scan A 654a258784fc
refine is a command published in the GitHub repository Connectry-io/connectrylab-architect-cert-mcp (33 stars, last pushed 5mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 513 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 2026-08-30.
Other commands, from other repositories
explore
Command "explore" from aderegil/claude-certified-architect, covering context: fork and allowed-tools: read, grep, glob.
review
Review the file at $ARGUMENTS for code quality issues.
review-issue
Review and respond to a GitHub issue.
review-pr-ci
Review a pull request and post the review to GitHub (CI/automated use).
review-pr
Review an open pull request and optionally post the review to GitHub.
add-registry
Add a new notebook to registry.yaml.