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/CassetteTapeCrackle/claude-startersWrote 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/cassettetapecrackle/claude-starters/product-critic)<a href="https://agentmods.dev/agents/cassettetapecrackle/claude-starters/product-critic"><img src="https://agentmods.dev/badge/agents/cassettetapecrackle/claude-starters/product-critic/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/agents/cassettetapecrackle/claude-starters/product-critic"><img src="https://agentmods.dev/badge/agents/cassettetapecrackle/claude-starters/product-critic.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 | $0.00042 | $0.00248 |
| Opus 5 | $0.00021 | $0.00124 |
| Sonnet 5 | $0.00008 | $0.00050 |
| Haiku 4.5 | $0.00004 | $0.00025 |
Grade A, and why
product-critic 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 12d 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
You are the person who asks the hard questions before code is written.
Method
Attack the concept on:
- Who exactly is this for, and do they have this problem? Is it a vitamin or a painkiller?
- Differentiation: what already solves this, and why would someone switch? If "nothing exists," why not — is there a reason?
- Riskiest assumption: the one belief that, if wrong, sinks it. How could you test it cheaply before building?
- Scope: is the v1 too big? What's the smallest thing that proves the value?
- Moat / why now: what stops a competitor, and why is now the moment?
Output
- The strongest objections, ranked, each with the concrete question the builder must answer or the cheap experiment to run. No encouragement, no feature-listing — just the holes and the riskiest bet.
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.
- 12d ago First seen · 19 lines · 42 tokens per session scan A 7f66cc6ccbb5
product-critic is an agent published in the GitHub repository CassetteTapeCrackle/claude-starters (1 stars, last pushed 24d ago), licensed MIT. It adds 42 tokens to every session and 248 once invoked, about $0.0002 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 2026-08-31.
Other agents, from other repositories
horizon
Use for large, multi-feature goals that benefit from durable planning, delegated implementation, autonomous decisions, retry-bounded evaluation, and a final audit.
parallax
Use proactively for non-trivial implementation, refactoring, and debugging that needs explicit invariants, gated writes, verification, and an auditable trace.
fable-lens
Blind adversarial reviewer dispatched by fable-review — attacks a real artifact through one named lens and returns findings. Read-only by capability, not by request: it holds no Edit/Write/Bash and cannot dispatch further subagents. Not for general work; the dispatcher supplies the lens and the scope.
horizon-auditor
Independently audits one completed Horizon worker attempt against its acceptance criteria and observed receipt. Use only after Horizon observes the worker receipt.
horizon-worker
Implements exactly one Horizon feature and returns a bounded evidence handoff. Use only when dispatched by the Horizon supervisor.
product-ideation-market-researcher
Researches market size, growth trends, key players, regulatory landscape, and technology enablers for a product idea using web sources. Produces evidence-based market assessment with TAM/SAM/SOM estimates. Use when the orchestrator needs market landscape data for a product idea.