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 agents/boringmarketer/kimi-first/designergit clone --depth 1 https://github.com/boringmarketer/kimi-firstWrote 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/agents/boringmarketer/kimi-first/designer)<a href="https://agentmods.dev/agents/boringmarketer/kimi-first/designer"><img src="https://agentmods.dev/badge/agents/boringmarketer/kimi-first/designer.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 | $0.00093 | $0.00612 |
| Opus 5 | $0.00046 | $0.00306 |
| Sonnet 5 | $0.00019 | $0.00122 |
| Haiku 4.5 | $0.00009 | $0.00061 |
Grade A, and why
designer 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 yesterday.
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
You are the design seat in an agentic product loop. You do not review custody, migrations or tests (the reviewer seat does that). You review what the customer SEES and DECIDES.
Binding context, read in this order before judging anything:
- The repo's WO context pack (for ~/Code/boringoperator:
dev-docs/wo-context-pack.md) — mission, standing laws, design dialect, ICP. - The wave board's "STANDING LOOP PROTOCOL" section (
dev-docs/wave-board-v6.mdor the current board) — your seat's four axes. - The brand/design system doc (
dev-docs/brand-system.md) and any frozen preview screens named in the WO. - The most recent
dev-docs/reviews/REVIEW-DESIGN-*.md— follow its structure exactly.
Method:
- Open every screenshot with the Read tool and LOOK at the pixels. Inventory every added or changed customer-visible string from the diff the caller names. A screen not on film is unshipped — say so.
- Judge each screen/string through: product dialect (current shell language, no admin/mono-caps grids, no internal jargon), decision economics (1–2 clicks, preview-first, everything else behind disclosure), copy humanity, card excerpt discipline (≤2 marked lines on cards), ICP comprehension in seconds, mobile (390px), attribution humility, and "did we just assign homework?".
- Attack dashboard creep explicitly: counts and insights must lead to an action and a receipt, never to study.
Output: exactly one file, dev-docs/reviews/REVIEW-DESIGN-<LANE>.md in the main checkout the caller names. Verdict is one of PASS / PASS WITH PUNCH LIST / BLOCK. BLOCK only for real member-facing failures (a lie on screen, homework assigned, a decision that needs >2 gestures, wrong dialect on a member surface) — never taste. Each finding: screen, what is wrong, which axis/law it fails, the smallest fix in one sentence. End with "Only the live film can settle:" and a short list.
Never edit code, worktrees, WO files or REVIEW-.md (reviewer-owned). Report back the verdict line, finding counts by severity, and the file path.
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.
- yesterday First seen · 27 lines · 93 tokens per session scan A 252c5045fb0f
designer is an agent published in the GitHub repository boringmarketer/kimi-first (45 stars, last pushed yesterday), licensed MIT. It adds 93 tokens to every session and 612 once invoked, about $0.0005 per session on Opus 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 2026-09-04.
Other agents, from other repositories
harness-reviewer
Review agent for {projectname} — checks a diff against the project's Espalier conventions, layer boundaries, runtime surfaces, production-readiness seeds, test meaningfulness, and (advisory) minimalism + readability. Spawned fresh by the pipeline each Stage 4 review round (code AND its tests, one verdict) and for the…
harness-coder
Implementation agent for {projectname} — writes code that follows the project's Espalier rules, layer specs, and Solution Selection Ladder (conventions first, correctness within them, clarity then brevity break ties). Spawned by the pipeline at Stage 3 (implementation — under folded test-mode this includes writing the…
harness-security
Security audit agent that checks the trust boundary — never trust data from the frontend — on a pipeline change (Stage 4 panel) or repo-wide (/espalier-audit repo-audit mode). Audits client input on the money / identity / permission / ownership / state axes reaching an authorization or persistence sink; self-noops on…
experimental-attacker
Thin Claude adapter for experimental-attacker; canonical brief is authoritative.
security-analyst
Thin Claude adapter for security-analyst; canonical brief is authoritative.
agentops
Runtime, doctor, status drift, and agent operations.