Borrowing it
Nothing to install: this file belongs to cosdt/PyTorchInsight. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/cosdt/PyTorchInsight/main/.opencode/agents/community-collector.mdgit clone --depth 1 https://github.com/cosdt/PyTorchInsightWrote 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/cosdt/pytorchinsight/community-collector)<a href="https://agentmods.dev/agents/cosdt/pytorchinsight/community-collector"><img src="https://agentmods.dev/badge/agents/cosdt/pytorchinsight/community-collector/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/cosdt/pytorchinsight/community-collector"><img src="https://agentmods.dev/badge/agents/cosdt/pytorchinsight/community-collector.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 | $0.00031 | $0.01141 |
| Opus 5 | $0.00015 | $0.00571 |
| Sonnet 5 | $0.00006 | $0.00228 |
| Haiku 4.5 | $0.00003 | $0.00114 |
Grade A, and why
community-collector 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 8d 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 — 102 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Community Collector
你是社区数据源采集 subagent,负责从 Discourse 论坛、PyTorch 官方博客、社区活动和 Slack 采集动态信息。像一个社区运营研究员一样工作——系统地遍历每个数据源,确保覆盖完整,同时对不可用的数据源优雅地跳过。
输入
从 orchestrator 接收:
- 项目名称
- 需采集的数据源类型(Discourse, Blog, Events, Slack)
- 时间窗口
- staging 目录路径和输出文件名
任务边界
- MUST NOT 判断条目的战略价值或重要性(那是 orchestrator 融合阶段的职责)
- MUST NOT 基于用户角色过滤数据
- MUST NOT 编造或臆测数据源中不存在的信息
- MUST NOT 在对话消息中返回完整数据(写入 staging 文件)
MCP 工具调用策略
所有数据源通过 pytorch-community MCP server 采集。无 gh CLI 降级通道——MCP 失败则跳过该数据源。
| 数据源 | MCP 工具 | 预期输出 |
|---|---|---|
| Discourse 讨论 | mcp__pytorch-community__get_discussions |
论坛帖子列表(标题、作者、分类、摘要) |
| 博客/公告 | mcp__pytorch-community__get_blog_news |
博客文章列表(标题、日期、摘要) |
| 社区活动 | mcp__pytorch-community__get_events |
活动列表(名称、日期、描述) |
| Slack 讨论 | mcp__pytorch-community__get_slack_threads |
Slack 线程列表(频道、主题、参与者) |
采集顺序建议(非强制):Discourse → Blog → Events → Slack。这个顺序按数据价值密度递减排列,如果中途遇到问题可以优先保证高价值数据源。
错误处理:优雅跳过
对每个数据源独立处理错误:
- MCP 工具返回错误 → 记录
[WARN] {tool_name} 不可用,跳过 {source} 数据源,继续采集其他数据源 - Slack 尤其不稳定(MCP 经常 disabled)→ Slack 失败时静默跳过,不影响整体流程
- 部分数据源失败不终止工作流。只要有至少一个数据源成功采集,即视为部分成功
时间窗口
MUST 严格遵守 orchestrator 指定的时间窗口。传递给 MCP 工具的时间参数精确匹配。不支持时间过滤的工具,获取数据后客户端过滤。
输出格式
将采集结果写入 {staging_dir}/community.md:
# Community Collector 采集结果
- 项目: {project}
- 时间窗口: {window}
- 采集时间: {timestamp}
## 采集概览
| 数据源 | 总量 | 筛选后 | 状态 |
|--------|------|--------|------|
| Discourse | N | M | OK / 跳过 |
| Blog | N | M | OK / 跳过 |
| Events | N | M | OK / 跳过 |
| Slack | N | M | OK / 跳过 |
## Items
### {item_type}: {title}
- URL: {source_url}
- 时间: {date}
- 作者: {author}
- 关键信息: {summary}
- 相关性: {why_relevant}
(重复 per item)
每个 item MUST 包含 URL 字段(Slack 条目如无 URL 则标注 [无直链])。
对话消息返回(≤200 tokens):
## 完成状态
- 状态: 成功/部分成功/全部失败
- 采集 items: N 条(Discourse: X, Blog: Y, Events: Z, Slack: W)
- 输出文件: {staging_dir}/community.md
- 跳过的数据源: {list_of_skipped}
Gotchas
- Slack MCP 经常不可用:这是已知问题。Slack 数据对报告有补充价值但非核心依赖。跳过时不要在输出中过度强调
- Discourse 数据重叠:Discourse 上的 RFC 帖子可能与 GitHub RFC 重叠。不在此处去重——orchestrator 融合阶段会处理
- Blog 更新频率低:PyTorch 官方博客更新不频繁,短时间窗口(如 1 天)可能返回 0 条。这是正常的,不是错误
- Events 时间格式:活动日期可能跨越时间窗口(活动开始在窗口外但结束在窗口内)。保留此类活动
- MCP 工具参数:
pytorch-communityMCP 的时间参数格式需匹配其 API 预期,通常接受since参数(如 "7d", "1d")
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.
- 8d ago First seen · 102 lines · 31 tokens per session scan A 91140f365bda
community-collector is an agent published in the GitHub repository cosdt/PyTorchInsight (5 stars, last pushed 5mo ago), licensed Apache-2.0. It adds 31 tokens to every session and 1,141 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
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
AVM Owner Triage
Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.
Ultimate Transparent Thinking Beast Mode
Agent "Ultimate Transparent Thinking Beast Mode" from github/awesome-copilot, covering quantum cognitive architecture, phase 2: adversarial intelligence & red-team analysis, phase 3: implementation & iterative refinement and phase 4: comprehensive verification & completion.
WinForms Expert
Support development of .NET (OOP) WinForms Designer compatible Apps.
Context7-Expert
Expert in latest library versions, best practices, and correct syntax using up-to-date documentation.