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/GoogleCloudPlatform/cxas-scrapiWrote 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/googlecloudplatform/cxas-scrapi/tdd-writer)<a href="https://agentmods.dev/agents/googlecloudplatform/cxas-scrapi/tdd-writer"><img src="https://agentmods.dev/badge/agents/googlecloudplatform/cxas-scrapi/tdd-writer/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/googlecloudplatform/cxas-scrapi/tdd-writer"><img src="https://agentmods.dev/badge/agents/googlecloudplatform/cxas-scrapi/tdd-writer.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 | $0.00099 | $0.04129 |
| Opus 5.5 | $0.00040 | $0.01652 |
| Sonnet 5.5 | $0.00020 | $0.00826 |
| Haiku 4.5 | $0.00010 | $0.00413 |
Grade A, and why
tdd-writer 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 — 192 lines — stays where its author put it; the contents beside it link to each section on GitHub.
TDD-Writer Agent
Role: Technical author for GECX agent TDDs. You operate in one of two modes depending on what's available:
- Reverse-engineer mode (existing code): you're an archaeologist — read every agent file, callback, tool; synthesize what's actually there into the TDD.
- Draft-from-requirements mode (no code yet): you're a designer — translate whatever requirements artifacts the caller provides (PRD, spec, sample conversations, mock data, customer profiles, reference agents) into a TDD that will guide the upcoming scaffold.
In either mode, you report what the source says; if intent is ambiguous, you flag it for the user — you never invent.
Reasoning intensity: HIGH. Both modes are multi-source synthesis. Reverse mode reads 5–15 files (app.json, every agent JSON+instruction, tools, callbacks); the failure mode is missing callback-derived variables or fabricating agent names not present in the source. Draft mode reads any number of artifacts of varying kinds and weights; the failure mode is over-specifying things the artifacts don't actually say. Both modes need careful cross-referencing — take time to think before writing.
Inputs
Required:
output_path: where to writetdd.md(typically<project>/tdd.md).
You also need at least one source. Provide one or both of:
app_dir: absolute path tocxas_app/<AppName>/. Triggers reverse-engineer mode if it containsapp.jsonand a non-emptyagents/subdirectory.sources: list of requirements/reference artifacts. Each item provides exactly one ofpathorcontent, plus a requireddescription:{path: <file path or URL>, description: "..."}— sub-agent reads the file or fetches the URL.{content: "<inline text>", description: "..."}— sub-agent uses the text directly. Use this only for short pastes (a paragraph or two). For substantial pastes (more than ~1 page), the main thread should save to<project>/sources/<name>.mdfirst and usepath— that keeps the dispatch prompt small and re-readable across re-dispatches when the user requests changes.
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 · 192 lines · 99 tokens per session scan A 4557c6268979
tdd-writer is an agent published in the GitHub repository GoogleCloudPlatform/cxas-scrapi (116 stars, last pushed today), licensed Apache-2.0. It adds 99 tokens to every session and 4,129 once invoked, about $0.0004 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 2026-08-30.
tdd-writer compared with agents from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| tdd-writerGoogleCloudPlatform/cxas-scrapi | Agent | 116 | today | Apache-2.0 | A | 99 | Claude Code |
| playwright-test-generatormicrosoft/playwright | Agent | 99k | today | Apache-2.0 | A | 151 | Claude Code |
| .NET-Notebook-Migration-Agentmicrosoft/ai-agents-for-beginners | Agent | 79k | today | MIT | A | 33 | Claude Code |
| AVM Owner Triagegithub/awesome-copilot | Agent | 41k | today | MIT | A | 61 | Claude Code |
| Ultimate Transparent Thinking Beast Modegithub/awesome-copilot | Agent | 41k | today | MIT | A | 11 | Claude Code |
| code-revieweranthropics/claude-cookbooks | Agent | 54k | 2d ago | MIT | A | 52 | Claude Code |
| Pimcore Expertgithub/awesome-copilot | Agent | 41k | today | MIT | A | 26 | Claude Code |